WO2022176945A1 - 生成プロセス出力装置、生成プロセス出力方法、及びプログラム - Google Patents

生成プロセス出力装置、生成プロセス出力方法、及びプログラム Download PDF

Info

Publication number
WO2022176945A1
WO2022176945A1 PCT/JP2022/006381 JP2022006381W WO2022176945A1 WO 2022176945 A1 WO2022176945 A1 WO 2022176945A1 JP 2022006381 W JP2022006381 W JP 2022006381W WO 2022176945 A1 WO2022176945 A1 WO 2022176945A1
Authority
WO
WIPO (PCT)
Prior art keywords
information
generation
unit
production
image
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Ceased
Application number
PCT/JP2022/006381
Other languages
English (en)
French (fr)
Japanese (ja)
Inventor
令子 羽川
洋正 玉置
麻由 森脇
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Panasonic Intellectual Property Management Co Ltd
Original Assignee
Panasonic Intellectual Property Management Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Panasonic Intellectual Property Management Co Ltd filed Critical Panasonic Intellectual Property Management Co Ltd
Priority to JP2023500920A priority Critical patent/JPWO2022176945A1/ja
Priority to CN202280015567.9A priority patent/CN116868220A/zh
Priority to EP22756263.4A priority patent/EP4296933A4/en
Publication of WO2022176945A1 publication Critical patent/WO2022176945A1/ja
Priority to US18/234,113 priority patent/US12494271B2/en
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16C—COMPUTATIONAL CHEMISTRY; CHEMOINFORMATICS; COMPUTATIONAL MATERIALS SCIENCE
    • G16C60/00—Computational materials science, i.e. ICT specially adapted for investigating the physical or chemical properties of materials or phenomena associated with their design, synthesis, processing, characterisation or utilisation
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/90—Details of database functions independent of the retrieved data types
    • G06F16/904—Browsing; Visualisation therefor
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/04—Manufacturing
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16C—COMPUTATIONAL CHEMISTRY; CHEMOINFORMATICS; COMPUTATIONAL MATERIALS SCIENCE
    • G16C20/00—Chemoinformatics, i.e. ICT specially adapted for the handling of physicochemical or structural data of chemical particles, elements, compounds or mixtures
    • G16C20/70—Machine learning, data mining or chemometrics
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
    • Y02P10/00—Technologies related to metal processing
    • Y02P10/25—Process efficiency

Definitions

  • the present disclosure relates to a production process output device and the like that outputs information about the production process of materials.
  • a synthesis process is a process for producing a target material by synthesis, and is also called a production process.
  • Patent Document 1 does not provide sufficient support for searching for the production process of the target material.
  • the present disclosure solves the above problems and provides a production process output device and the like that can more appropriately support the search for the production process of the target material.
  • a generation process output device includes a first information acquisition unit that acquires first information about a target material, and a target material specified by the first information for each of one or more types of generation processes. a derivation unit for deriving, based on a database of materials, a production parameter indicative of the degree to which a production process of that type is required for production; and an output unit for outputting at least one of said derived production parameters.
  • FIG. 1 is a diagram showing an example of the configuration of a generated process search system according to each embodiment.
  • FIG. 2 is a block diagram showing an example of the functional configuration of the generating process display device according to each embodiment.
  • FIG. 3 is a diagram showing an example of a prediction result of the baking method by the baking predictor of the baking learning unit.
  • FIG. 4 is a block diagram showing an example of the functional configuration of the database construction device according to each embodiment.
  • FIG. 5 is a flowchart showing an example of the overall flow of processing of the generated process search system according to each embodiment.
  • FIG. 6 is a flowchart showing an example of the overall flow of processing by the database construction device according to each embodiment.
  • FIG. 1 is a diagram showing an example of the configuration of a generated process search system according to each embodiment.
  • FIG. 2 is a block diagram showing an example of the functional configuration of the generating process display device according to each embodiment.
  • FIG. 3 is a diagram showing an example of a prediction result of the baking method by the baking
  • FIG. 7 is a flowchart showing an example of the flow of pre-processing for learning in the combined information learning unit according to each embodiment.
  • 8A is a diagram showing an example of a process-related image according to Embodiment 1.
  • FIG. 8B is a diagram showing another example of the process-related image according to Embodiment 1.
  • FIG. 8C is a diagram showing another example of the process-related image according to Embodiment 1.
  • FIG. 8D is a diagram showing an example of a reference image according to Embodiment 1.
  • FIG. 8E is a diagram showing another example of a reference image according to Embodiment 1.
  • FIG. 9 is a flowchart illustrating an example of the overall flow of processing of the generated process display device according to Embodiment 1.
  • FIG. 9 is a flowchart illustrating an example of the overall flow of processing of the generated process display device according to Embodiment 1.
  • FIG. 9 is a flowchart illustrating an example of the overall flow of processing of the generated process display
  • FIG. 10A is a diagram showing an example of a process-related image according to Embodiment 2.
  • FIG. 10B is a diagram showing another example of a process-related image according to Embodiment 2.
  • FIG. 10C is a diagram showing an example of a map according to Embodiment 2.
  • FIG. 10D is a diagram showing another example of a map according to Embodiment 2.
  • FIG. 10E is a diagram showing another example of a map according to Embodiment 2.
  • FIG. FIG. 11 is a flowchart showing an example of the overall flow of processing of the generated process display device according to the second embodiment.
  • 12A is a diagram showing an example of a process-related image according to Embodiment 3.
  • FIG. 12B is a diagram illustrating an example of a process-related image after updating according to Embodiment 3.
  • FIG. 13 is a flowchart illustrating an example of the overall flow of processing of the generated process display device according to Embodiment 3.
  • FIG. 14 is a diagram showing another example of the process-related image according to Embodiment 3.
  • FIG. 15A is a diagram showing an example of a process-related image according to Embodiment 4.
  • FIG. 15B is a diagram illustrating an example of a process-related image after updating according to Embodiment 4.
  • FIG. 16 is a flowchart illustrating an example of the overall flow of processing of the generated process display device according to Embodiment 4.
  • FIG. 17A is a diagram showing an example of an initial image according to Embodiment 5.
  • FIG. 17B is a diagram showing an example of a process-related image according to Embodiment 5.
  • FIG. 17C is a diagram showing another example of a process-related image according to Embodiment 5.
  • FIG. 18 is a flowchart illustrating an example of the overall flow of processing of the generated process display device according to Embodiment 5.
  • FIG. 19 is a diagram showing an example of the configuration of a generated process search system according to Embodiment 6.
  • FIG. 20 is a block diagram illustrating an example of a functional configuration of a generation process display device according to Embodiment 6.
  • FIG. FIG. 21 is a block diagram showing an example of the functional configuration of the database construction device according to the sixth embodiment.
  • FIG. 22A is a diagram showing an example of information indicated by a second database according to Embodiment 6.
  • FIG. 22B is a diagram showing an example of information indicated by a second database according to Embodiment 6.
  • FIG. 23 is a flowchart illustrating an example of the overall flow of processing of the generated process search system according to Embodiment 6.
  • FIG. 24 is a flow chart showing an example of the overall flow of processing by the database construction device according to the sixth embodiment.
  • 25 is a diagram illustrating an example of a process-related image according to Embodiment 6.
  • FIG. 26 is a diagram illustrating another example of a process-related image according to Embodiment 6.
  • FIG. 27 is a flowchart illustrating an example of the overall flow of processing of the generated process display device according to Embodiment 6.
  • FIG. 28 is a diagram illustrating an example of a process-related image according to Modification 1 of Embodiment 6.
  • FIG. 29 is a diagram illustrating another example of a process-related image according to Modification 1 of Embodiment 6.
  • FIG. 30 is a diagram illustrating another example of a process-related image according to Modification 1 of Embodiment 6.
  • FIG. 31 is a diagram showing an example of a process-related image Pa1 according to Modification 2 of Embodiment 6.
  • FIG. 32 is a diagram showing an example of a process-related image Pa1 according to Modification 2 of Embodiment 6.
  • FIG. 33 is a diagram illustrating another example of a process-related image according to Modification 2 of Embodiment 6.
  • FIG. 34 is a diagram illustrating an example of a process-related image according to Modification 3 of Embodiment 6.
  • FIG. 35 is a diagram illustrating another example of a process-related image according to Modification 3 of Embodiment 6.
  • FIG. 35 is
  • a generation process output device includes a first information acquisition unit that acquires first information about a target material, and for each of one or more types of generation processes, the above a deriving unit for deriving, based on a material database, a production parameter indicating the degree to which the type of production process is required to produce the target material specified by the first information; and at least one derived production parameter. and an output unit that outputs
  • the output unit outputs the generation parameter indicating the degree as at least one of a generation probability and a rank, and the generation probability indicates that a type of generation process corresponding to the generation probability is used to generate the target material.
  • said rank may be the rank in an array in which each of said one or more generation processes is ordered in order of said degree.
  • generation of materials is, for example, synthesis of materials
  • synthesis of materials is, for example, firing of materials.
  • the one or more production processes may also be, for example, a calcination method, sonication, or the like.
  • the production parameters for each type of production process for the target material are output. That is, for each type of generation process, the user can easily grasp the necessity of the generation process for the generation of the target material, in other words, the possibility of generating the target material by the generation process. Therefore, when generating a target material, the user can select a generation process after grasping the possibility of generation for each type of generation process. As a result, even if the user is a researcher with little knowledge and experience, he or she can easily select the generation process. Also, if the user is a researcher with a wealth of knowledge and experience, the user's knowledge and experience can be used to select a more accurate generation process.
  • the user finds the type of generation process that is most required for the generation of the target material from those generation processes. be able to. That type of production process can also be said to be the type of production process most likely to produce the desired material. The user can then determine that the production process can be used to produce the target material. As a result, the user's search for the production process of the target material can be more appropriately supported.
  • the derivation unit outputs a generation parameter of each of the one or more types of generation processes for the material in response to the input of the descriptor related to the material, to the predictor learned using the database.
  • the production parameters for each of the one or more production processes for the target material may be derived by inputting descriptors for the target material.
  • the predictor is a neural network or the like.
  • the predictor uses the database so that, for input of a descriptor indicating a composition, generation parameters of each of the one or more types of generation processes for a material having the composition are output.
  • the derivation unit inputs a descriptor indicating the composition of the target material to the predictor, so that the generation of each of the one or more types of generation processes for the target material parameters may be derived.
  • the predictor outputs a production parameter of each of the one or more production processes for the material for input of descriptors indicating one or more raw materials for producing the material. It is a predictor trained using a database, and the derivation unit inputs descriptors indicating one or more raw materials for generating the target material into the predictor so that the one or more raw materials for the target material are input.
  • the production parameters for each of one or more types of production processes may be derived.
  • the production parameters of each of the one or more production processes for the unknown target material can be obtained by using the learned predictor. can be derived.
  • the output unit may output the generation parameters by generating a first image showing at least one of the generation parameters and outputting the first image to a display unit.
  • the first image that is, the process-related image
  • the user can generate the target material using the type of generation process corresponding to the largest generation parameter. can be determined.
  • the user confirms the generation parameters corresponding to the generation processes possible with the generation device that is owned. be able to.
  • the user may try to generate the target material by the generation process even if the generation process corresponds to the smallest generation parameter, if the generation parameter indicates a generation probability of 30% or more, for example. judgment can be made.
  • the production process output device further comprises a second information acquisition unit that acquires second information indicating the known material for each of the one or more known materials related to the target material from a material database,
  • the output unit generates the first image further indicative of the second information for each of the one or more known materials, and the second information acquisition unit further comprises the one or more known materials indicated by the first image.
  • the second information indicating the known material and the partial area of the document describing the production process of the known material are displayed.
  • the user can select a production process after grasping the possibility of production for each type of production process from a broader point of view when producing a target material.
  • the user can easily grasp what kind of production process is being performed for a known material similar to the target material. Therefore, it is possible to more appropriately support the search for the production process of the target material.
  • the generation parameter of each of N types of generation processes (N is an integer of 3 or more) is derived by the derivation unit, and the degree of When the generation parameter of each of the M types of generation processes (M is a predetermined integer of 1 or more less than N) satisfies a predetermined condition, the output unit outputs, among the N types of generation processes,
  • the first image may be generated showing the production parameters of each of the M kinds of production processes and not showing the production parameters of the remaining other production processes.
  • the predetermined condition is that the generation probability is equal to or greater than the threshold value of 80%. Therefore, the user can easily find the most likely production process for the target material. Further, for example, if the sum of the generation probabilities of the top two of the generation probabilities of three or more types of generation processes is equal to or greater than the threshold value of 90%, only the top two generation probabilities are displayed, and the remaining other generation probabilities are displayed.
  • the top two generation probabilities are two generation probabilities that are higher than any other generation probabilities.
  • the predetermined condition is that the sum of the top two generation probabilities is equal to or greater than the threshold value of 90%. Therefore, it is possible to omit from the first image production parameters such as production probability indicating that there is no possibility of production of the target material, and it is possible to suppress useless provision of information to the user.
  • the second information acquisition unit acquires the second information indicating the composition formula and attributes of the known material
  • the output unit further acquires (i) the second information of each of the one or more known materials (ii) a map showing the relationship between the composition formula and the attribute indicated by the second information of each of the one or more known materials; or (iii) the A map showing the relationship of the attributes indicated by the second information of each of the one or more known materials may be generated and output to the display unit.
  • the second information is displayed on the map, so that the user can easily find the desired second information from the second information. be able to.
  • the user can easily select the desired second information and easily view the second image (that is, the reference image) corresponding to the second information.
  • the attributes of the known material indicated by the second information include the crystal structure of the known material, process conditions in the production process for producing the known material, characteristic values indicating the degree of characteristics of the known material, At least one of literature information for identifying literature describing the known material and use of the known material may be provided.
  • the process conditions are conditions used in the production process, such as temperature and time.
  • the user can know the attributes of known materials, and can easily guess information that will serve as a reference for the production of the target material based on those attributes.
  • the output unit further specifies a type of generation process that can be executed by the generation device owned by the user, and selects the specified type from among the generation parameters of each of the one or more types of generation processes. of each of the one or more types of updated generation processes by weighting differently the degree indicated by the generation parameter of the generation process and the degree indicated by the remaining generation parameters
  • the first image may be generated indicative of the generated parameters.
  • the second information acquisition unit further provides an estimate of a generation device that can be executed by the remaining generation processes excluding the specified type of generation process among the one or more types of generation processes.
  • the output unit may further output an image showing the estimate information to the display unit.
  • the user can easily request an estimate for a generator that the user does not own, according to the displayed estimate information.
  • the second information of each of the one or more known materials includes an estimated required time that is the time required to generate the known material
  • the output unit outputs the second information of each of the one or more known materials may be arranged in order according to the estimated required time indicated by the second information, and the first image showing one or more of the second information arranged in the order may be generated.
  • the estimated required time for each of one or more known materials is displayed, so the user can predict the time required to produce the target material based on the estimated required times.
  • the estimated required times are arranged and displayed in order, the user can easily grasp the maximum value, minimum value, or variance of the estimated required times, which is necessary for generating the target material. Time can be predicted better.
  • the second information of each of the one or more known materials also indicates a production process for producing the known material, any production process shortens the time required to produce the target material.
  • the user can easily determine whether the For example, even if the user intends to use the generation process with the largest generation parameter shown in the first image to generate the target material, if the estimated time required for the generation process is long, the next largest generation parameter may be generated. It can be determined that the process is used to produce the target material. In other words, the user can select the production process of the target material in consideration of the time required for production.
  • each figure is a schematic diagram and is not necessarily strictly illustrated. Moreover, in each figure, the same code
  • Embodiments 1 to 5 Prior to describing the details of the first to fifth embodiments, the configuration of the generative process search system 10 commonly used in the first to fifth embodiments will be described. Incidentally, in Embodiments 1 to 5, the firing method in particular among the production processes will be described.
  • the material development of inorganic materials also called inorganic compounds
  • the firing method is generally a method of changing properties by firing raw materials at a high temperature, and various types of firing methods are known. The optimum firing method largely depends on the combination and composition ratio of elemental species possessed by the inorganic material.
  • firing method is also called a sintering method.
  • firing of materials is a subordinate concept of synthesis of materials, and synthesis of materials is a subordinate concept of production of materials. As such, calcination may be referred to as synthesis or generation.
  • a means for predicting a route (that is, a synthesis route) for synthesizing a target molecule from a compound available on the market by applying reaction rules accumulated in a reaction knowledge base, and each predicted synthesis
  • a method for estimating the variable cost index which is a rough estimate of the cost required to produce a target molecule from starting materials, for each route, and presenting the predicted multiple synthetic routes to the user in ascending order of the estimated variable cost index.
  • a synthetic route design system comprising means is disclosed. With this synthetic route design system, users can easily determine which synthetic route is economically and industrially feasible among many predicted synthetic routes.
  • Patent Document 1 that is, the system that supports the search for synthesis routes, does not have the function of displaying the baking method that is likely to synthesize the target material, which is the target molecule. Therefore, when synthesizing the target material, it is difficult for the user to set the experimental conditions after grasping the possibility of synthesis for each firing method.
  • Embodiments 1 to 5 a generation process search system 10 that allows a user to easily grasp the possibility of synthesis for each firing method will be described.
  • FIG. 1 is a diagram showing an example of the configuration of a generated process search system 10 according to Embodiments 1-5.
  • the generated process search system 10 includes a generated process display device 100 , a display unit 108 , an input unit 110 , a database construction device 200 , a first storage unit 301 and a second storage unit 302 .
  • the generation process display device 100, the database construction device 200, the first storage unit 301, and the second storage unit 302 are connected to each other via a network 401 such as the Internet.
  • the production process display device 100 is configured as a computer such as a personal computer or a server, and displays information on the production process, which is the process for producing the target material, on the display unit 108 .
  • the material is produced by synthesizing the material. Therefore, material production is also called material synthesis. Therefore, the production process is also called a synthesis process. More specifically, production or synthesis of the material involves calcination of the material. Therefore, the production process or synthesis process in Embodiments 1 to 5 means the method of firing the material.
  • the display unit 108 displays an image showing information about the above-described generation process, that is, the synthesis process, according to the signal output from the generation process display device 100 .
  • a display unit 108 is, for example, a liquid crystal display, a plasma display, an organic EL (Electro-Luminescence) display, or the like, but is not limited to these.
  • the input unit 110 is configured as, for example, a keyboard, a touch sensor, a touch pad, a mouse, or the like, receives an input operation by the user of the generating process display device 100, and outputs a signal corresponding to the input operation to the generating process display device 100.
  • the display unit 108 and the input unit 110 are configured independently of each other in Embodiments 1 to 5, they may be configured integrally like a touch panel. Further, in Embodiments 1 to 5, the generating process display device 100 does not include the display unit 108 and the input unit 110, but may include them.
  • the database construction device 200 is configured as, for example, a computer such as a personal computer or a server, and constructs the second database used in the generation process display device 100 .
  • the database construction device 200 reads the first database stored in the first storage unit 301 from the first storage unit 301 via the network 401, and uses the first database to construct the second database. .
  • the database construction device 200 stores the constructed second database in the second storage unit 302 via the network 401 .
  • the first storage unit 301 is a recording medium for storing the first database.
  • the first database includes a plurality of article data each indicating the content of an article on material synthesis.
  • the second storage unit 302 is a recording medium for storing the second database. Both the first database and the second database are databases related to materials. Details of the second database will be described later.
  • These recording media are, for example, hard disk drives, RAMs (Random Access Memory), ROMs (Read Only Memory), or semiconductor memories. Note that such a recording medium may be volatile or nonvolatile.
  • first storage unit 301 and the second storage unit 302 are arranged outside the database construction device 200 in the example shown in FIG. Also, the first storage unit 301 and the second storage unit 302 may be directly connected to the database construction device 200 without going through the network 401 . Also, the first database and the second database are stored in different recording media, but may be stored in the same recording medium. Furthermore, the set of the first database and the second database may be treated as one database.
  • FIG. 2 is a block diagram showing an example of the functional configuration of the generating process display device 100 according to Embodiments 1-5.
  • the generated process display device 100 uses a first database stored in the first storage unit 301 and a second database stored in the second storage unit 302 .
  • the first database contains a plurality of article data 1 .
  • Each of the plurality of article data 1 is associated with the composition formula of the main material to be synthesized (that is, the known material) described in the article data 1 .
  • the second database includes a plurality of first combined information 2a and a plurality of second combined information 2b.
  • Each of the plurality of pieces of first synthesis information 2a shows the composition formula of the known material associated with the article data 1 and the type name of the firing method of the known material in association with each other.
  • Each of the plurality of second combined information 2b indicates the composition formula of the known material associated with the article data 1 and the attributes of the known material indicated in the article data 1 in association with each other.
  • the attribute of the known material may be the literature information of the article data 1 indicating the known material, or the location of the paragraph describing the synthesis of the known material in the article data 1, or the like.
  • the attributes of the known material may be the crystal structure, sintering temperature, characteristic value, application, or the like of the known material.
  • the attributes of the known material may include all or part of each of the above items.
  • the document information is information for identifying the article data 1, and is, for example, all or part of the article title, issue date, publication year, and author name indicated by the article data 1.
  • the location of a paragraph is, for example, the paragraph number, title, or item number of that paragraph in the article data 1 .
  • the second database in Embodiments 1 to 5 shows, for each of a plurality of materials, the compositional formula, firing method, and attributes of the material in association with each other.
  • acquisition of the second information and derivation of the generation probability can be performed appropriately, as will be described later.
  • the data format of the second database that is, the data format of the first combined information 2a and the second combined information 2b may be, for example, a tabular format or a json (JavaScript Object Notation) format.
  • the production process display device 100 is a device for displaying information about the production process of materials.
  • the production process display device 100 displays the production probability for each type of firing method for the target composition formula.
  • the target composition formula is the composition formula of the material to be synthesized (that is, the target material).
  • the generation probability for each type of firing method for the target material or target composition formula is the probability that the target material having the target composition formula is fired by the firing method for each of a plurality of predetermined firing methods. is the probability of indicating For example, the generation probability is expressed as a percentage. The higher the production probability of the firing method for the target material or target composition formula, the higher the probability of firing the target material by that firing method.
  • the generation process display device 100 is connected to the first input section 101 and the second input section 109 included in the input section 110 and the display section 108 .
  • the generation process display device 100 includes a first information acquisition unit 102, a composition descriptor generation unit 103, a firing learning unit 104, a probability derivation unit 105, a second information acquisition unit 106, and an image generation unit 107.
  • the generating process display device 100 does not include the first input section 101, the second input section 109, and the display section 108, but does include at least one of them. may
  • the first input unit 101 is a functional component included in the input unit 110, and by receiving an input operation of the user, for example, the composition formula according to the input operation, that is, the purpose of the target material to be displayed A signal indicating the compositional formula is output to the first information acquisition unit 102 .
  • First information acquisition section 102 receives a signal from first input section 101 . That is, the first information acquisition unit 102 acquires first information, which is a signal indicating the target composition formula. First information acquisition section 102 outputs the first information to composition descriptor generation section 103 and image generation section 107 .
  • composition descriptor generator 103 acquires the first information from the first information acquisition unit 102 and generates a composition descriptor corresponding to the target composition formula indicated by the first information.
  • a composition descriptor is, for example, a vector that uniquely indicates a composition formula. Specifically, when a composition formula consists of a combination of one or more element species out of 72 element species, the composition descriptor corresponding to the composition formula is a 72-dimensional vector. For example, in the case of the composition formula “CaMnO 3 ”, the composition descriptor corresponding to the composition formula “CaMnO 3 ” is indicated by a three-dimensional vector consisting of a combination of one or more element species among the three element species.
  • composition descriptor [1, 1, 3] may be normalized so that the vector sum is 1 and represented as [0.2, 0.2, 0.6].
  • the composition descriptor is a weighted average, maximum value or minimum value of known parameters unique to each element.
  • the known parameters peculiar to each element refer to a group of known numerical values possessed by each element, such as atomic volume, covalent bond radius or density, which can be obtained without physical calculation.
  • a weighted average of the parameters is calculated based on the number of atoms that make up the material.
  • the composition descriptor generation unit 103 outputs the composition descriptor to the probability derivation unit 105 and the second information acquisition unit 106 .
  • the firing learning unit 104 refers to the second database of the second storage unit 302 and learns the relationship between the composition formula and the firing method, thereby constructing a firing predictor that indicates the relationship.
  • the firing learning unit 104 uses a plurality of pieces of first combined information 2a included in the second database as teacher data.
  • Each of the plurality of pieces of first synthesis information 2a shows a composition formula and a type name of a firing method for a known material having the composition formula in association with each other.
  • the firing learning unit 104 outputs the composition formula indicated in the first synthesis information 2 a to the composition descriptor generation unit 103 and acquires the composition descriptor corresponding to the composition formula from the composition descriptor generation unit 103 .
  • the firing learning unit 104 determines the composition descriptor corresponding to the composition formula indicated in the first synthesis information 2a and the firing method indicated in the first synthesis information 2a. Learning is performed using the type name as teacher data. As a result, the firing learning unit 104 constructs a firing predictor that outputs the generation probability of each of the plurality of types of firing methods in response to the input of the composition descriptor.
  • the generation probability of each of the multiple types of firing methods is a probability indicating the possibility that the material having the composition formula corresponding to the input composition descriptor is fired by that type of firing method, that is, synthesized.
  • the firing learning unit 104 acquires the composition descriptor corresponding to the target composition formula from the probability derivation unit 105, the firing learning unit 104 inputs the composition descriptor to the firing predictor, thereby obtaining a plurality of types of firing methods for the target composition formula. is obtained from the firing predictor. Firing learning section 104 then outputs these generation probabilities to probability deriving section 105 .
  • the firing learning unit 104 re-learns the firing predictor based on the difference between the post-update and the pre-update.
  • FIG. 3 is a diagram showing an example of the prediction result of the baking method by the baking predictor of the baking learning unit 104.
  • FIG. The map shown in FIG. 3 is a two-dimensional map having a vertical axis and a horizontal axis.
  • Li, B, C, O, F, Na, Mg, Al, Si, P, S, Cl, K, Ca, Sc, Ti, V, Cr, Mn, Fe, Co, Ni, Cu, Zn, Ga, Ge, As, Se, Br, Rb, Sr, Y, Zr, Nb, Mo, Ru, Rh, Pd, Ag, Cd, In, Sn, Sb, Te, I, Cs, Ba, La, Ce, Pr, Nd, Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb, Lu, Hf, Ta, W, Re, Ir, Pt, Au, Hg, Tl, Pb and Bi are employed.
  • the plurality of elemental species described above are arranged in the order described above so that the intersection of the vertical axis and the horizontal axis (that is, the lower left end in FIG. 3) is the end point.
  • the above-described plurality of element species are arranged in the above-described order so that their intersection point is the starting point.
  • the firing learning unit 104 uses the plurality of first combined information 2a acquired from the second storage unit 302 and the multiple composition descriptors acquired from the composition descriptor generation unit 103 to learn the firing predictor.
  • the types of firing methods in Embodiments 1 to 5 are Ball-Milling, Solid-State, Liquid-State, Flux, and Arc-Melting.
  • the firing predictor of the firing learning unit 104 outputs the generation probability of each of the five types of firing methods with respect to the input of the composition descriptor corresponding to the compositional formula of the material.
  • the two-dimensional map of FIG. 3 shows the composition formula corresponding to the composition descriptor input to the firing predictor, and the production probability of each of the five types of firing methods output from the firing predictor for that composition formula. Among them, the type of firing method corresponding to the highest probability of generation is shown in association.
  • the composition formula is a composition formula in which one element type on the vertical axis and one element type on the horizontal axis are combined at a ratio of 1:1.
  • the production probability When one of the combined element species is Li and the other element species is C, Ti, Rb, Y, Zr, Mo, Ru, Ce, Gd, Er, Tm, or Tl, the production probability
  • the two-dimensional map shows that the firing method with the highest ⁇ is Arc-Melting.
  • the other element species is F, Rh, Nd, Ta, Re, Ir, or Pt
  • the two-dimensional map shows that the firing method with the highest generation probability is Solid-State.
  • the other element species is Cl, Br, Tb, Ho, Lu, W, or Au
  • the two-dimensional map shows that the firing method with the highest generation probability is Liquid-State.
  • the two-dimensional map also shows that when the other element species is Na or Ba, the firing method with the highest generation probability is Flux.
  • the two-dimensional map shows that when one of the combined element species is Rb, the firing method with the highest generation probability is Arc-Melting, regardless of which element the other element species is. .
  • the type of element to be learned and the type of firing method are examples, and are not limited to these.
  • the arrangement order of the element species is also an example, and may be arranged with emphasis on the ease of correspondence between the element species on the vertical axis and the horizontal axis, or may be arranged with emphasis on the visibility of the two-dimensional map.
  • the probability derivation unit 105 acquires the composition descriptor corresponding to the target composition formula from the composition descriptor generation unit 103, and generates the probability that the target material having the target composition formula is generated by each of the five firing methods. Derived as a probability.
  • a firing predictor of the firing learning unit 104 is used for deriving the generation probability. That is, the probability derivation unit 105 calculates the probability that the target material having the target composition formula indicated in the first information will be fired by that type of firing method for each of one or more types of firing methods, as the generation probability. It is derived based on the second database in the storage unit 302 .
  • the probability derivation unit 105 inputs the composition descriptor acquired from the composition descriptor generation unit 103 to the firing predictor of the firing learning unit 104 . Then, the probability derivation unit 105 acquires the generation probability of each of the five types of baking methods output from the baking predictor from the baking predictor. The probability derivation unit 105 outputs the generation probability of each of the five types of baking methods to the second information acquisition unit 106 and the image generation unit 107 .
  • the composition descriptor generation unit 103 For example, if the target composition formula is BiCuSO, the composition descriptor generation unit 103 generates a composition descriptor that is a vector representing BiCuSO. Then, the probability derivation unit 105 acquires the composition descriptor from the composition descriptor generation unit 103, and inputs the composition descriptor to the learned firing predictor of the firing learning unit 104 to determine the firing method for the BiCuSO. Derive the generation probability for each type of
  • the probability derivation unit 105 derives the generation probability based on the combination and composition ratio of the element species included in the target composition formula indicated by the first information. This makes it possible to increase the reliability of the derived generation probability.
  • the probability deriving unit 105 in response to the input of the composition descriptor indicating the composition formula, outputs the generation probability of each of one or more firing methods for the material having the composition formula.
  • the generation probability of each of one or more types of firing methods for the target material is derived.
  • the learned firing predictor is used to derive the generation probability of each of one or more firing methods for the unknown target material. be able to.
  • the second information acquisition unit 106 acquires the composition descriptor of the target composition formula from the composition descriptor generation unit 103 and acquires the generation probability of each of the five types of baking methods from the probability derivation unit 105 . Then, the second information acquiring unit 106 acquires one or more pieces of second information necessary for image generation based on the acquired composition descriptors and the generation probabilities of each of the five types of firing methods. from the second database of Further, second information acquisition section 106 outputs the second information to image generation section 107 . As a result, an image representing one or more pieces of second information (that is, a process-related image to be described later) is generated by the image generator 107 and displayed.
  • the second information is, for example, a composition formula of a known material that is predicted to have the same firing method as the target composition formula input by the user and that is associated with the target composition formula, and is associated with the composition formula. It is information that indicates the document information and attributes such as the location of the paragraph.
  • the same firing method for the target composition formula and the known material composition formula means, for example, that the target material and the known material have the same type of firing method corresponding to the highest generation probability. do.
  • the second information is not limited to the above example.
  • the second information may indicate attributes such as crystal structure, sintering temperature, property values, and application corresponding to the compositional formula of the known material.
  • the second information acquiring unit 106 obtains a first synthesis information indicating a composition formula of a known material related to the target composition formula among the plurality of first synthesis information 2a and the plurality of second synthesis information 2b included in the second database.
  • the second information is obtained by obtaining the information 2a and the second synthesized information 2b. Note that the firing method indicated by the acquired second information is treated as an attribute of the known material.
  • the second information acquisition unit 106 retrieves the article corresponding to the selected second information.
  • the paragraph of data 1 is obtained from the first database and output to the image generator 107 .
  • the second information acquisition unit 106 acquires the composition formula and attributes of each of one or more known materials related to the target composition formula indicated in the first information.
  • the second information shown is acquired from the second database of the second storage unit 302 .
  • the attributes of the known material indicated by the second information include the crystal structure of the known material, the firing temperature for firing the known material, the characteristic value indicating the degree of the properties of the known material, and the known material. and/or the use of the known material.
  • the document is article data 1 in the first database. Attributes may also be the location of paragraphs in the literature describing known materials and/or firing methods.
  • the second information obtaining unit 106 selects the second information indicated by the selected second information.
  • a part of a document describing a method of firing a known material having a composition formula is acquired from the first database of the first storage unit 301 .
  • the image generation unit 107 acquires the target composition formula from the first information acquisition unit 102 and generates an image showing the target composition formula. In addition, the image generator 107 generates an image showing the generation probabilities acquired from the probability derivation unit 105 . In other words, the image generation unit 107 generates an image showing the generation probability of each of the five firing methods corresponding to the composition descriptor of the target composition formula. Also, the image generation unit 107 acquires one or more pieces of second information from the second information acquisition unit 106 and generates an image showing the second information.
  • the second information in Embodiments 1 to 5 indicates the compositional formula, crystal structure, firing temperature, characteristic values, literature information, application, etc., but is not limited to this.
  • Each generated image described above is included in, for example, a process-related image and displayed on the display unit 108 .
  • the image generation unit 107 uses the image representing the second information of each of the one or more known materials and the generation probability of each of the one or more firing methods as a process-related image representing information on the generation process. It is generated and displayed on the display unit 108 . Further, as described above, when a partial area in the document is acquired by the second information acquisition unit 106, the image generation unit 107 generates an image showing the description content of the partial area in the document. A reference image is generated and displayed on the display unit 108 . Note that the reference image is also called a second image. Also, some areas in the document are, for example, paragraphs, as described above.
  • the display unit 108 acquires a signal from the image generation unit 107 and displays an image according to the acquired signal. That is, the display unit 108 displays the process-related image and the reference image described above.
  • the second input section 109 is a functional component included in the input section 110 .
  • the second input unit 109 selects, for example, any one of the one or more pieces of second information in the process-related image displayed on the display unit 108 by accepting a user's input operation.
  • Second input section 109 then outputs a signal for identifying the selected second information to second information acquiring section 106 .
  • the second information acquisition unit 106 identifies the second information selected by the user from among the one or more pieces of second information, and describes the baking method corresponding to the selected second information. Get the paragraph in the article data 1.
  • FIG. 4 is a block diagram showing an example of the functional configuration of the database construction device 200 according to Embodiments 1-5.
  • the database construction device 200 analyzes the information in each article data 1 from the first database collected in advance, and the second data including the first synthetic information 2a showing the composition formula and the corresponding firing method in association with each other. Build a database.
  • This second database also contains the above-described second synthesized information 2b.
  • Information in the constructed second database is acquired by the second information acquisition unit 106 of the production process display device 100 and used by the baking learning unit 104 .
  • the database construction device 200 selects, from all paragraphs in the article data 1 to which the composition formula is assigned in advance, the paragraph describing the synthesis of the material of the composition formula. identify.
  • the material described in the paper data 1 is a known material.
  • the database construction device 200 acquires the name, which is the type name of the firing method, in the specified paragraph, and associates the name of the firing method with the composition formula given to the article data 1. generates the first combined information 2a and adds it to the second database.
  • the database construction device 200 uses a pre-learned synthesis predictor to identify paragraphs that describe the synthesis of materials. Further, the database construction device 200 extracts information used by the image generation unit 107 by analyzing the article data 1 . For example, the database construction device 200 detects and extracts paragraphs describing the firing method from the article data 1 .
  • Such a database construction device 200 includes an article acquisition unit 202, an area division unit 203, an area descriptor generation unit 204, a labeling unit 205, a synthetic information learning unit 206, a label acquisition unit 207, and an information extraction unit. 208;
  • the article acquisition unit 202 acquires the article data 1 from the first database of the first storage unit 301 and outputs it to the region dividing unit 203 .
  • the region division unit 203 extracts the text information of the paper data 1 acquired from the paper acquisition unit 202, and divides the text information into a plurality of regions. Each of these multiple areas is, for example, a paragraph. That is, the region division unit 203 divides the text information into a plurality of paragraphs and acquires them. For example, ScienceParse is used for extracting text information from the article data 1 of Embodiments 1 to 5, but it is not limited to this. Then, the region division unit 203 outputs each acquired paragraph to the label acquisition unit 207 , the region descriptor generation unit 204 and the label assignment unit 205 .
  • Region descriptor generating section 204 generates region descriptors for each of the paragraphs acquired from region dividing section 203 , and outputs the region descriptors to synthesizing information learning section 206 .
  • a region descriptor is, for example, a vector that uniquely indicates what is written in a paragraph.
  • the label assigning unit 205 assigns labels to the paragraphs obtained from the area dividing unit 203 according to the input operation by the user of the database construction device 200 . This label indicates whether the paragraph contains information about synthesizing the material. The label assigning unit 205 then outputs the assigned label to the combined information learning unit 206 and the label acquiring unit 207 .
  • Synthetic information learning unit 206 acquires region descriptors corresponding to each of the plurality of paragraphs from region descriptor generation unit 204 and acquires labels assigned to each of the plurality of paragraphs from labeling unit 205 . Synthetic information learning section 206 then learns the relationship between the region descriptor corresponding to each of the paragraphs and the label attached to each of the paragraphs, thereby creating a synthetic predictor that indicates the relationship. To construct. For example, clustering is used as the prediction means. Clustering is a type of unsupervised learning in machine learning, and is a method of grouping data based on similarity between data. Techniques include, but are not limited to, k-means, random forests, neural networks, and the like. The composition predictor, given a region descriptor input, outputs a label indicating whether the paragraph corresponding to the region descriptor contains information regarding the composition of the material.
  • the synthetic information learning unit 206 acquires the region descriptor from the region descriptor generating unit 204 without acquiring the label from the labeling unit 205 .
  • the synthesis information learning unit 206 inputs the region descriptor to the synthesis predictor, thereby obtaining the above-described label corresponding to the region descriptor from the synthesis predictor.
  • Synthetic information learning section 206 outputs the label acquired from the synthetic predictor to label acquiring section 207 .
  • the label acquisition unit 207 acquires a paragraph from the region division unit 203 and acquires a label corresponding to the paragraph from the label assignment unit 205 or the synthetic information learning unit 206 .
  • the label obtaining unit 207 obtains a label from the label assigning unit 205 .
  • the label acquisition unit 207 acquires the label from the synthetic information learning unit 206 .
  • the label obtaining unit 207 gives the paragraph obtained from the region dividing unit 203 a label corresponding to the paragraph obtained from the label assigning unit 205 or the combined information learning unit 206 .
  • the label acquisition unit 207 then outputs the labeled paragraph to the information extraction unit 208 .
  • the information extraction unit 208 acquires the labeled paragraph from the label acquisition unit 207 .
  • Information extractor 208 then extracts the information from the paragraphs labeled as containing information about the synthesis of ingredients.
  • the information is information about the production process of the material, and indicates, for example, at least one of the name of the sintering method, the composition formula, the sintering temperature, the crystal structure, and the characteristic values.
  • the information extraction unit 208 extracts information indicating the name of the baking method from the paragraph by searching for a character string indicating the name of the baking method from the paragraph.
  • the information extracting unit 208 extracts information indicating the firing temperature from the paragraph by searching the unit of the firing temperature and the numerical value placed before the unit from the paragraph.
  • the information extracting unit 208 extracts information indicating the crystal structure from the paragraph by searching for a character string indicating the crystal structure from the paragraph. Further, the information extracting unit 208 extracts information indicating the characteristic value from the paragraph by searching the unit of the characteristic value and the numerical value placed before the unit from the paragraph.
  • the information extraction unit 208 associates the firing method indicated by the extracted information with the composition formula given in advance to the paper data 1 acquired by the paper acquisition unit 202, thereby obtaining the composition formula of the known material and the firing method.
  • First combined information 2a is generated that indicates the method in association with the method.
  • the information extraction unit 208 associates the composition formula of the known material with the attribute by associating the attribute of the known material other than the firing method with the composition formula of the article data 1 acquired by the article acquisition unit 202.
  • the second synthesized information 2b shown is generated.
  • This attribute may include the document information of the article data 1 acquired by the article acquisition unit 202 and the location of the paragraph from which the information indicating the burning method was extracted.
  • the attributes may include the firing temperature, crystal structure, characteristic values, etc. indicated by the information extracted by the information extractor 208 .
  • FIG. 5 is a flow chart showing an example of the overall flow of processing of the generated process searching system 10 according to the first to fifth embodiments.
  • Step S1100 The database construction device 200 constructs a second database including first synthesis information 2a and second synthesis information 2b that indicate the compositional formulas and firing methods of known materials in association with each other.
  • Step S1200 The generation process display device 100 constructs a firing predictor that predicts the firing method from the composition formula by performing learning using the first synthesis information 2a.
  • a fully-connected NeuralNetwork for example, is used as the firing predictor model.
  • Step S1300 The production process display device 100 uses a sintering predictor to derive the production probability for each type of sintering method for the target composition formula. In other words, the production process display device 100 predicts the firing method for firing the target material having the target composition formula.
  • FIG. 6 is a flow chart showing an example of the overall flow of processing by the database construction device 200 according to the first to fifth embodiments. That is, it is a flowchart showing the detailed flow of step S1100.
  • Embodiments 1 to 5 for example, 1063 article data 1 are used. Each of these article data 1 is associated with the composition formula of the thermoelectric conversion material. The association between the article data 1 and the composition formula is performed by manually confirming the composition formula mainly handled in the article data 1 and assigning it to the article data 1 .
  • the method of assigning the compositional formula is not limited to this, and may be automatic.
  • Step S1101 The paper acquisition unit 202 acquires one paper data 1 to be processed from the first database containing a plurality of paper data 1 already published, and outputs the paper data 1 to the area division unit 203 .
  • the article data 1 to which the compositional formula is linked in advance is selected.
  • Step S1102 The region division unit 203 divides the article data 1 acquired from the article acquisition unit 202 into a plurality of paragraphs and acquires them.
  • the region division unit 203 outputs each obtained paragraph to the label acquisition unit 207 , the region descriptor generation unit 204 , and the label assignment unit 205 .
  • Step S1103 Upon obtaining a paragraph from the region dividing unit 203, the region descriptor generating unit 204 generates a region descriptor corresponding to the paragraph.
  • the synthesis information learning unit 206 inputs the region descriptor to the synthesis predictor to determine whether or not the paragraph corresponding to the region descriptor contains information on material synthesis. Gets a label that indicates whether Combined information learning section 206 then outputs the label to label obtaining section 207 .
  • the label assigning unit 205 outputs the above label to the label acquiring unit 207 according to the input operation by the user of the database constructing apparatus 200 .
  • the label acquisition unit 207 acquires the label output from the synthesis information learning unit 206 or the labeling unit 205, that is, the label indicating whether or not the paragraph describes the synthesis of ingredients, and assigns the acquired label to the paragraph. do.
  • the label acquisition unit 207 then outputs the labeled paragraph to the information extraction unit 208 .
  • Step S1104 The information extraction unit 208 extracts information indicating the name of the baking method, etc. from the paragraphs describing the synthesis of the materials, that is, the paragraphs containing the information on the synthesis of the materials, among the paragraphs acquired from the label acquisition unit 207. do.
  • a plurality of names of baking methods are registered in advance in the information extraction unit 208 .
  • information extraction unit 208 determines whether or not the name of the registered firing method exists in the paragraph describing synthesis of materials. When the information extracting unit 208 determines that there is a name of the registered baking method, the information extracting unit 208 extracts the name of the baking method.
  • the information extracting unit 208 extracts all the baking method names.
  • the information extracted by the information extraction unit 208 indicates not only the name of the firing method but also attributes other than the firing method related to the composition formula of the article data 1 acquired in step S1101. This attribute is the document information of the paper data 1, the location of the paragraph that describes the name of the firing method, or the crystal structure, firing temperature, characteristic value, or use of the composition formula described in the paragraph and so on.
  • Step S1105) The information extraction unit 208 further associates the name of the firing method extracted in step S1104 with the composition formula given to the article data 1 having the paragraph describing the name of the firing method, Add to database. That is, the first combined information 2a is generated and added to the second database. Further, the information extraction unit 208 associates the attribute other than the firing method extracted in step S1104 with the composition formula given to the article data 1, and adds them to the second database. That is, the second combined information 2b is generated and added to the second database.
  • FIG. 7 is a flowchart showing an example of the flow of preprocessing for learning in combined information learning section 206 according to Embodiments 1-5. That is, FIG. 7 shows part of the processing included in the processing of step S1103 in FIG.
  • the area descriptor generation unit 204 generates an area descriptor expressed as a vector by converting the character string included in the paragraph acquired from the area division unit 203 into a vector using Word2Vec, for example.
  • Synthesis information learning section 206 inputs a vector, which is the region descriptor, to a trained synthesis predictor to determine whether the paragraph corresponding to the region descriptor is a paragraph describing synthesis of materials. Get a label that indicates the .
  • the Word2Vec model is a specialized model for materials papers published by the Synthesis Project. In Embodiments 1 to 5, Word2Vec is used for conversion from character strings of paragraphs to vectors, but other models may be used without being limited to this.
  • Machine learning is used for learning of the synthetic predictor by the synthetic information learning unit 206 .
  • This machine learning model uses stochastic gradient descent classification.
  • a learning database is used for learning of the synthetic predictor.
  • the training database contains each paragraph of each of the 100 article data 1, labeled as to whether or not the paragraph is about synthesis of materials.
  • the flowchart shown in FIG. 7 is a flowchart for manually generating the learning database.
  • Step S1111 The paper acquisition unit 202 acquires one paper data 1 from the first database and outputs it to the area division unit 203 . That is, one article data 1 is selected.
  • some article data 1 included in the first database is used, but the present invention is not limited to this. Included article data may be used.
  • 100 articles of article data 1 are randomly selected from the first database for learning in the combined information learning unit 206, and for each of the 100 articles of article data 1, the data shown in FIG. Each process shown in the flowchart is repeatedly executed.
  • Step S1112 The region division unit 203 divides the article data 1 acquired from the article acquisition unit 202 into a plurality of paragraphs and acquires them.
  • the region division unit 203 outputs each obtained paragraph to the label acquisition unit 207 , the region descriptor generation unit 204 , and the label assignment unit 205 .
  • Step S1113 After acquiring a paragraph from the region dividing unit 203, the labeling unit 205 mechanically pre-screens the paragraph. For example, if the paragraph starts with a word such as Summary or refer that has nothing to do with material synthesis (Yes in step S1113), the labeling unit 205 performs the process in step S1117 and then the process in step S1118. conduct. On the other hand, if the paragraph does not start with a word unrelated to material synthesis (No in step S1113), the labeling unit 205 performs the process of step S1114.
  • a word such as Summary or refer that has nothing to do with material synthesis
  • Step S1114 it is manually checked whether or not a paragraph determined not to start with a word unrelated to material synthesis contains information about material synthesis. That is, the labeling unit 205 accepts an input operation by the user of the database construction device 200, and determines whether or not the paragraph contains information regarding the synthesis of materials according to the input operation.
  • Step S1115) If the paragraph contains information on material synthesis (Yes in step S1115), the labeling unit 205 performs the process of step S1116, and if the paragraph does not contain information on material synthesis (step S1115 No), the process of step S1117 is performed.
  • Step S1116 The label assigning unit 205 assigns a label of 1, ie, a label indicating that information regarding synthesis of materials is included, to the paragraph confirmed in step S1114.
  • Step S1117) The labeling unit 205 assigns a label of 0, that is, a label indicating that no information about material synthesis is included, to the paragraph confirmed in step S1114 or a paragraph starting with a word unrelated to material synthesis. Give.
  • Step S1118 The combined information learning unit 206 adds the label given by the label giving unit 205 and the area descriptor of the paragraph corresponding to the label to the learning database. That is, the label is associated with the region descriptor of the paragraph and added to the training database.
  • a learning database is constructed by repeating the processing of steps S1111 to S1118, for example, 100 times.
  • the database construction device 200 builds the above-described learning database, and further, while learning the synthesis information learning unit 206 using the learning database, the first synthesis Information 2a and second combined information 2b are generated.
  • the database construction device 200 assigns a label to each paragraph using the synthetic predictor instead of the labeling unit 205. Attachment is performed to generate the first combined information 2a and the second combined information 2b.
  • FIG. 1 An image displayed by the generating process display device 100 according to the first embodiment will be described.
  • 8A to 8E are diagrams showing an example of display on the generating process display device 100 according to Embodiment 1.
  • FIG. 1 An image displayed by the generating process display device 100 according to the first embodiment will be described.
  • FIG. 8A is a diagram showing an example of a process-related image according to Embodiment 1.
  • the image generation unit 107 generates a process-related image P1 and displays it on the display unit 108.
  • FIG. This process-related image P1 includes a target composition region d1, a known material region d2, and a production probability region d3.
  • a target composition formula indicated by the first information acquired by the first information acquisition unit 102 is displayed in the target composition area d1.
  • One or more pieces of second information acquired by the second information acquisition unit 106 are displayed in the known material area d2.
  • the generation probability area d3 displays the generation probability of each of the five firing methods for firing the target material having the target composition formula, which is derived by the probability derivation unit 105.
  • the target composition formula "BiCuSO" input by the user is displayed.
  • BiCuSO and Ag 3 Si can be said to be compositional formulas representing the matrix, which is a structural element that occupies most of the volume of the alloy, but is not limited to this.
  • the target composition formula may be a real composition formula including dopant species, such as Bi 0.9 Mg 0.1 CuSO. Note that the actual composition formula is also called the composition formula of the actual composition.
  • the generation probability for each type of firing method for the target composition formula "BiCuSO” is displayed in a table d31 and a pie chart d32.
  • the types of firing methods in Embodiments 1 to 5 are Ball-Milling, Solid-State, Liquid-State, Flux, and Arc-Melting, but are not limited to these.
  • the displayed firing method type is based on the first synthesis information 2a.
  • sub-areas having tabs are displayed for each type of firing method.
  • a sub-region displays second information for each of one or more known materials related to the target material to be fired by the type of firing method corresponding to that sub-region.
  • the second information indicates the compositional formula and attributes of the known material, and in the example of FIG. 8A, the attributes are the firing method and literature information.
  • the image generation unit 107 in the present embodiment generates the process-related image P1 indicating the second information of each of the one or more known materials generated by the type of firing method for each type of firing method. do.
  • This allows the user to easily find out from the process-related image P1 the second information of each of the one or more known materials fired by the firing method with the highest production probability for the target material, for example. That is, even if a large amount of second information is displayed, the user can easily find useful second information regarding firing of the target material.
  • a sub-area having a tab for the type of firing method "Ball-Milling” is displayed first.
  • This firing method type "Ball-Milling” is the type of firing method corresponding to the highest generation probability among the five generation probabilities derived for the target composition formula "BiCuSO” among the five types of firing methods. is. That is, the image generation unit 107 identifies the type of firing method corresponding to the highest generation probability among the five generation probabilities derived for the target composition formula, and tabs indicating the identified firing method type is first displayed on display 108 . Furthermore, the image generator 107 may arrange the five tabs in descending order of generation probability. This allows the user to easily visually understand the firing method related to the target material.
  • the composition formula and attributes of each of one or more known materials related to the target material are displayed as second information.
  • Known materials that are related to the target material are materials that are similar to the target material and have close vector distances to the target material. That is, known materials that are related to the target material are materials that have composition descriptors that are close to the target material's composition descriptor.
  • the composition formulas of one or more known materials each having a vector-to-vector distance equal to or less than a threshold from the composition descriptor of the target material are displayed in the known material area d2 in descending order of the distance. Note that a predetermined number of composition formulas of known materials may be displayed in the known material area d2 in ascending order of distance.
  • the image generation unit 107 in the present embodiment arranges the second information of each of the one or more known materials in the order according to the composition formula indicated by the second information, and arranges the second information in the order.
  • a process-related image P1 indicating one or more pieces of second information is generated.
  • the second information indicating the composition formula of a known material that is close to or similar to the composition formula of the target material is arranged first. Thereby, the user can easily find out the desired second information from the second information.
  • composition formulas "BiCuSeO” and “BiCuTeO” of known materials which are composition formulas related to the target composition formula "BiCuSO” and have the highest probability of producing Ball-Milling, are displayed.
  • elemental species Se and Te that differ from the target compositional formula are highlighted in bold.
  • the composition formula of the known material related to the target composition formula may be the real composition formula including the dopant species, or may be displayed as the composition formula of the base material in the known material region d2 for easy understanding. good.
  • the compositional formula of the base material is a compositional formula that does not contain dopant species such as BiCuSeO.
  • the composition formula of the actual composition may be displayed, or only the composition formula of the actual composition may be displayed.
  • the composition formula of the actual composition is a composition formula including dopant species.
  • the dopant species are Mg, Ca, Sr, and Ba
  • their composition formula is a composition formula in which the dopant species are contained at a ratio of 0 ⁇ x ⁇ 0.15.
  • the attribute of the known material displayed in the known material area d2 includes, for example, the document information of the paper data 1 that is the source of the composition formula of the known material.
  • Literature information that is the source of the composition formula is obtained from the second synthetic information 2b of the second database.
  • the literature information may be acquired from other databases via network 401, such as the Internet.
  • combination information 2b links
  • the second synthesis information 2b may indicate a main composition formula among the plurality of composition formulas.
  • FIG. 8B is a diagram showing another example of the process-related image P1 according to Embodiment 1.
  • FIG. 8B similarly to the process-related image P1 shown in FIG. 8A, the composition formula of the target material input by the user and the generation probability of the firing method related to the composition formula are displayed.
  • the target composition formula "BiRbSO" input by the user is displayed.
  • the production probability of the firing method for the target composition formula "BiRbSO” is displayed by a table d31 and a pie chart d32.
  • the target composition formula is "BiRbSO”
  • Arc-Melting only one type of firing method, Arc-Melting, is displayed in the generation probability area d3 among the five types of firing methods.
  • the production probability area d3 only the production probability of "85.0%” for Arc-Melting is displayed among the production probabilities for each of the five types of baking methods.
  • the image generation unit 107 determines the generation probability of 80% or more and the generation probability Only the name of the corresponding firing method is included in the production probability area d3. Then, the image generation unit 107 does not include the four generation probabilities other than the generation probability of 80% or more and the name of the baking method corresponding to each of the four generation probabilities in the generation probability area d3.
  • the generation probability of 80% is an example of a threshold, and the threshold may be higher or lower than 80%.
  • Such a generation probability is based on the learning result in the firing learning unit 104 as explained in FIG. That is, when one element species included in the target composition formula consisting of two element species is Rb, the formation probability is the highest among the five types of firing methods regardless of which element species the other element species is.
  • the firing learning unit 104 has learned that the firing method is Arc-Melting.
  • the known material area d2 only sub-areas with Arc-Melting tabs are displayed.
  • This subregion displays the formula and attributes of each of the one or more known materials associated with the target formula "BiRbSO".
  • Their compositional formulas are those of Rb-containing materials that are likely to be sintered by Arc-Melting.
  • the attribute is the document information that is the source of the compositional formula of the known material.
  • the image generation unit 107 when a predetermined condition is satisfied, indicates the generation probability of each of the M types of baking methods among the N types of baking methods, and the remaining other baking methods. Generate a process-related image P1 that does not show the generation probabilities of the method.
  • N is an integer of 3 or more
  • M is a predetermined integer of 1 or more less than N.
  • the M types of firing methods are one type of firing method, specifically Arc-Melting.
  • the above-mentioned predetermined conditions are such that the generation probability of each of the N types of firing methods is derived by the probability derivation unit 105, and among the N types of firing methods, the generation probability is higher than the remaining other firing methods.
  • the image generator 107 generates a process-related image P1 showing the top two generation probabilities.
  • the image generation unit 107 indicates the top two generation probabilities and the name of the baking method corresponding to each of the two generation probabilities, but the remaining three generation probabilities and each of the three generation probabilities. generates a process-related image P1 that does not indicate the name of the firing method corresponding to .
  • the generation probability of one type of firing method among a plurality of firing methods is a threshold value of 80% or more, only the generation probability of that one type of firing method is displayed, and the remaining firing methods are displayed. Generation probabilities are not displayed. Therefore, the user can easily find the most likely firing method for the target material.
  • the generation probabilities of each of the three or more types of baking methods for example, if the sum of the top two generation probabilities is a threshold value of 90% or more, only the top two generation probabilities are displayed, and the remaining other generation probabilities are displayed. Probability is not displayed. Therefore, it is possible to omit from the process-related image P1 the production probability that the target material is unlikely to be fired, and to suppress unnecessary information provision to the user.
  • the image generation unit 107 may display only the element species of Rb among the composition formulas of the one or more known materials displayed in the known material region d2, emphasizing them more than the other element species. Since it has been found by learning that the material having a composition formula including Rb has the highest generation probability of Arc-Melting, the image generation unit 107 emphasizes and displays the element species of Rb. For example, if the generation probability of any one of the generation probabilities of each of the five types of baking methods is equal to or greater than a threshold, the image generation unit 107 identifies the type of baking method corresponding to the generation probability equal to or greater than the threshold. do. In the example above, the threshold is 80%. Note that the image generation unit 107 may emphasize only element species that differ from the target composition formula among the composition formulas of the known materials.
  • FIG. 8C is a diagram showing another example of the process-related image P1 according to Embodiment 1.
  • FIG. 8C In the known material region d2 in the process-related image P1 shown in FIG. 8C, not only the document information shown in FIG. 8A but also the crystal structure corresponding to the composition formula of the known material, the firing Temperatures, property values, and applications are displayed.
  • the known material region d2 displayed in FIG. 8C is an example, and is not limited to this.
  • the second information displayed in the known material area d2 may be customized by the user.
  • the image generation unit 107 in the present embodiment arranges the second information of each of the one or more known materials in an order according to the attribute indicated by the second information, and arranges the one or more A process-related image P1 showing the second information may be generated. For example, if the attribute is the firing temperature of a known material, the second information indicating the higher firing temperature is arranged first. This allows the user to easily find the desired second information from the second information.
  • FIG. 8D is a diagram showing an example of a reference image according to Embodiment 1.
  • FIG. 8D For example, while the process-related image P1 shown in FIG. 8C is being displayed, the user performs an input operation on the second input unit 109, whereby the second information indicating the composition formula “BiCuSeO” in the known material region d2 is displayed. to select.
  • second input section 109 outputs a signal corresponding to the input operation to second information acquisition section 106 .
  • the second information acquisition unit 106 acquires the signal from the second input unit 109, and specifies "BiCuSeO", which is the composition formula of the second information identified by the signal.
  • the second information acquisition unit 106 searches for the second synthesized information 2b indicating the specified composition formula “BiCuSeO”, and finds the location of the document information and the paragraph indicated as attributes in the second synthesized information 2b. identify. Further, the second information acquisition unit 106 searches the first database of the first storage unit 301 for the paper data 1 having the specified document information, and from the paper data 1, the paragraph at the specified location is extracted, and the paragraph is output to the image generation unit 107 . The image generation unit 107 superimposes the reference image P2 indicating the paragraph acquired from the second information acquisition unit 106 on the process-related image P1 and displays it on the display unit 108 .
  • FIG. 8D when the composition formula "BiCuSeO" is selected by the user, a reference image P2 showing a paragraph describing the generation process of the article data 1 corresponding to the composition formula "BiCuSeO” is displayed as an example. . Further, the image generation unit 107 may emphasize the character string of the composition formula “BiCuSeO” and the character string of the firing method “Ball-Milling” in the paragraph more than other character strings. For example, the image generation unit 107 acquires the composition formula “BiCuSeO” specified by the second information acquisition unit 106 from the second information acquisition unit 106, and searches for the character string of the composition formula “BiCuSeO” from the paragraph.
  • the image generation unit 107 searches the paragraph for the character string of the firing method “Ball-Milling” corresponding to the sub-region where the composition formula “BiCuSeO” is indicated. Then, the image generation unit 107 emphasizes the searched character string of the composition formula “BiCuSeO” and the character string of the firing method “Ball-Milling”.
  • the image generation unit 107 in the present embodiment causes the word representing the firing method indicated as an attribute by the second information selected in the partial area of the document such as the article data 1 to be added to the remaining area.
  • a reference image P2 is generated in which the other words are emphasized.
  • the word representing the firing method is highlighted, so that the user can easily find out the word representing the firing method from a part of the article data 1 .
  • the user can search around the word to easily find the description related to the firing method.
  • the user can confirm at a glance the generation process described in thesis data 1 using the firing method predicted to have the highest generation probability for the selected composition formula "BiCuSeO". can be done. That is, in the present embodiment, first, the paragraph shown in the reference image P2 is the paragraph of the paper data 1 using the firing method predicted to have the highest generation probability, so for the target material Probable generation processes can be displayed to the user. Secondly, it is not necessary for the user to read or search for the relevant part of the generation process from the contents of the paper data 1, so the efficiency of the generation process search work itself can be improved.
  • FIG. 8E is a diagram showing another example of the reference image P2 according to Embodiment 1.
  • FIG. 8E For example, while the process-related image P1 shown in FIG. 8C is displayed, the user selects a plurality of composition formulas in the known material area d2 by performing an input operation on the second input unit 109. That is, a plurality of pieces of second information are selected.
  • second input section 109 outputs a signal corresponding to the input operation to second information acquiring section 106 .
  • the second information acquisition unit 106 acquires the signal from the second input unit 109 and identifies a plurality of composition formulas selected by the user indicated by the signal.
  • the second information acquiring unit 106 searches for the second combined information 2b indicating the composition formula for each of the specified plurality of composition formulas, and searches for the document indicated as the attribute in the second combined information 2b. Locate information and paragraphs. Further, the second information acquisition unit 106 searches the first database of the first storage unit 301 for the paper data 1 having the specified document information, and retrieves the paragraph at the above-mentioned specified location from the paper data 1. The paragraph is extracted and output to the image generation unit 107 . The second information acquisition unit 106 outputs the paragraph to the image generation unit 107 for each of the plurality of compositional formulas selected by the user.
  • the image generation unit 107 generates a reference image P2 showing each paragraph acquired from the second information acquisition unit 106 in a list format or a tab format, and superimposes it on the process-related image P1 and displays it on the display unit 108. .
  • FIG. 8E when the user selects three compositional formulas, a reference image P2 showing paragraphs corresponding to each of the three compositional formulas in a tab format is displayed.
  • This tab format display is an example, and the paragraphs corresponding to each composition formula may be displayed in another format, or may be displayed in a list format.
  • This can facilitate comparison of production processes. For example, a plurality of paragraphs of article data 1 using the firing method predicted to have the highest generation probability are selected and displayed in a list format.
  • the ease of comparing the production processes described in the selected paragraphs so that even if the user is an inexperienced researcher, using the firing method predicted to have the highest probability of production. It is possible to facilitate the examination of the production process that has a high possibility of realization for the target material.
  • the user's knowledge and experience will be used to compare the production processes and examine the production process that further increases the possibility of realization for the target material. be able to
  • FIG. 9 is a flowchart showing an example of the overall flow of processing of the generating process display device 100 according to Embodiment 1 of the present disclosure. 9 is a flowchart showing the detailed flow of step S1300 in FIG.
  • the first input unit 101 receives a target composition formula to be displayed by a user's input operation, and outputs the target composition formula to the first information acquisition unit 102 .
  • the first information acquisition unit 102 acquires the target composition formula from the first input unit 101 and outputs it to the composition descriptor generation unit 103 .
  • the composition descriptor generation unit 103 generates a composition descriptor corresponding to the target composition formula acquired from the first information acquisition unit 102 and outputs it to the probability derivation unit 105 and the second information acquisition unit 106 .
  • the firing learning unit 104 inputs the composition descriptors obtained from the composition descriptor generation unit 103 through the probability derivation unit 105 to the learned firing predictor, thereby learning the composition descriptors for each type of firing method corresponding to the composition descriptors. , and outputs the generation probability to the probability derivation unit 105 .
  • the baking learning unit 104 may re-learn the baking predictor based on the difference between the post-update and the pre-update.
  • the probability derivation unit 105 acquires the generation probability for each type of firing method from the firing learning unit 104, thereby deriving the generation probability for each type of firing method for the target composition formula.
  • the probability derivation unit 105 outputs the generation probability for each type of baking method to the image generation unit 107 .
  • the image generation unit 107 generates a process-related image P ⁇ b>1 obtained from the first information acquisition unit 102 and showing the target composition formula input by the user.
  • This process-related image P1 includes the generation probability for each type of firing method for the target composition formula, which is acquired from the probability derivation unit 105 .
  • this process-related image P1 includes second information of each of the one or more known materials related to the target material, which is acquired from the second information acquisition unit 106 .
  • the second information indicates, for example, the compositional formula of the known material and the attributes of the known material. Attributes include crystal structure, sintering temperature, sintering method, characteristic values, literature information, location of paragraphs, usage, and the like.
  • Step S1306 The image generation unit 107 displays the process-related image P1 generated in step S1305 on the display unit .
  • Step S1307) The second information acquisition unit 106 selects the second information according to the user's input operation to the second input unit 109 from among the one or more pieces of second information included in the process-related image P1.
  • Step S1308 The second information acquisition unit 106 acquires the paragraph corresponding to the composition formula from the first database based on the composition formula indicated by the selected second information, and outputs the paragraph to the image generation unit 107 .
  • the image generating unit 107 After obtaining the paragraph from the second information obtaining unit 106, the image generating unit 107 generates a reference image P2 indicating the paragraph.
  • Step S1309) The image generation unit 107 displays the generated reference image P2 on the display unit 108.
  • the production process display device 100 includes the first information acquisition unit 102 that acquires the first information indicating the composition formula of the target material, and For each of the above-described known materials, a second information acquisition unit 106 that acquires second information indicating the composition formula and attributes of the known material from a material database, and for each of one or more types of firing methods, the first information a probability deriving unit 105 for deriving the probability that a target material having a composition formula represented by is fired by that type of firing method as a generation probability based on a database; second information for each of one or more known materials; , and the generation probability of each of the one or more baking methods as a process-related image P1 representing information about the generation process, and displays the generated image on the display unit 108 .
  • the second information acquisition unit 106 has a composition formula indicated by the selected second information when one of the second information of each of the one or more known materials indicated by the process-related image P1 is selected. Some areas of the literature describing methods of firing known materials are retrieved from the database.
  • the image generation unit 107 generates an image showing the description content of a part of the document as the reference image P2 and displays it on the display unit 108 .
  • the database is, for example, at least one of the first database of the first storage unit 301 and the second database of the second storage unit 302 .
  • the document is the article data 1, for example.
  • the user confirms the generation probability corresponding to the firing method possible with the firing apparatus that is owned. be able to.
  • the user can try firing the target material by the firing method if the generation probability is, for example, 50% or more.
  • the composition formula and attributes of each of the one or more known materials related to the target material for example, the composition formula and attributes of each of the one or more known materials having a composition formula similar to the target material are displayed in the process-related image P1 can be shown in The attribute may be a matter of firing, synthesizing or producing a known material. As a result, the user can easily grasp what kind of firing is being performed on a known material similar to the target material. Therefore, it is possible to more appropriately support the search for the production process of the target material.
  • FIG. 1 An image displayed by the generating process display device 100 according to the second embodiment will be described.
  • the generated process display device 100 according to the second embodiment displays a map showing one or more pieces of second information in response to an input operation by the user. Furthermore, the generating process display device 100 switches the display form of the map.
  • 10A to 10E are diagrams showing examples of display on the generation process display device 100 according to the second embodiment.
  • the image generation unit 107 adds a map display button d22 and a similar composition formula to the process-related image P1 shown in FIG. , sintering temperature, characteristic values, and radio buttons d21 for each of the document information.
  • FIG. 10A is a diagram showing an example of the process-related image P1 when the similar composition formula radio button d21 is selected.
  • the one or more composition formulas in the known material region d2 that is, the composition formulas indicated by the one or more pieces of second information, are not arranged in descending order of similarity to the target composition formula, they are arranged in that order.
  • the display order of the one or more compositional formulas may be switched as shown in FIG.
  • the degree of similarity indicates a higher value as the distance between the vectors in the composition descriptor of the target composition formula and the composition descriptor of the similar composition formula is shorter.
  • FIG. 10B is a diagram showing an example of the process-related image P1 when the map display button d22 is further selected after the similar composition formula radio button d21 is selected in FIG. 10A.
  • FIG. 10C is a diagram showing an example of a map displayed after the map display button d22 in FIG. 10B is selected.
  • the image generation unit 107 displays a similar composition map m1 indicating the relationship between the composition formula of one or more known materials and the target composition formula superimposed on the process-related image P1.
  • the similar composition map m1 is a periodic table of elements.
  • the similar composition map m1 indicates, on the periodic table of elements, elements that are not included in the target composition formula among the composition formulas of one or more known materials related to the target material.
  • the similar composition map m1 indicates elements on the periodic table of elements that are not included in the composition formulas of the one or more known materials among the target composition formulas.
  • the second information acquisition unit 106 displays the paragraph corresponding to the composition formula including the selected element as the first 1 database and output to the image generation unit 107 .
  • the image generation unit 107 When acquiring the paragraph from the second information acquisition unit 106, the image generation unit 107 generates and displays a reference image P2 showing the paragraph.
  • FIG. 10D is a diagram showing an example of a map displayed after the firing temperature radio button d21 is selected and the map display button d22 is selected.
  • the image generation unit 107 displays a firing temperature map m2 indicating the relationship between firing temperatures of one or more known materials related to the target material, superimposed on the process-related image P1.
  • the firing temperature map m2 is a two-dimensional map
  • the horizontal axis of the firing temperature map m2 indicates a plurality of composition formulas
  • the vertical axis of the firing temperature map m2 indicates the firing temperature.
  • the composition formulas of the one or more known materials are arranged, for example, in order of closeness to the target composition formula.
  • the second information acquisition unit 106 selects a composition formula on the firing temperature map m2 or a dot indicating the firing temperature of the composition formula according to the user's input operation to the second input unit 109.
  • a paragraph corresponding to the formula or dot drawn is obtained from the first database.
  • Second information acquisition section 106 then outputs the paragraph to image generation section 107 .
  • the image generating unit 107 Upon obtaining the paragraph from the second information obtaining unit 106, the image generating unit 107 generates and displays a reference image P2 showing the paragraph.
  • FIG. 10E is a diagram showing an example of the map displayed after the radio button d21 of the document information is selected and the map display button d22 is selected.
  • the image generator 107 superimposes and displays a document map m3 indicating the relationship of the article data 1 corresponding to each of the plurality of known materials on the process-related image P1.
  • the horizontal axis of the document map m3 indicates the publication year of the article data 1
  • the vertical axis of the document map m3 indicates the name of the author of the article data 1.
  • the publication year and author name are shown as document information in the second information. For example, a user may know researchers who have been working on known materials for many years.
  • the second information acquisition unit 106 selects the dot corresponding to the selected dot.
  • the article data 1 to be used is specified from the first database.
  • the second information acquisition unit 106 outputs the paragraph of the paper data 1 that includes the firing method to the image generation unit 107 .
  • the image generating unit 107 Upon obtaining the paragraph from the second information obtaining unit 106, the image generating unit 107 generates and displays a reference image P2 showing the paragraph.
  • the image generation unit 107 in the present embodiment generates, for example, the similar composition map m1 as a map representing the relationship between the composition formulas indicated by the second information of each of the one or more known materials, and displays it on the display unit. 108. Further, the image generation unit 107 generates, for example, a firing temperature map m2 as a map showing the relationship between the composition formula and the attribute indicated by the second information of each of the one or more known materials, and displays it on the display unit 108. .
  • the second information can be displayed on the map corresponding to the radio button d21. . Therefore, the user can easily find the desired second information from those pieces of second information. As a result, the user can easily select the desired second information and easily view the reference image P2 corresponding to the second information.
  • FIG. 11 is a flowchart showing an example of the overall flow of processing of the generating process display device 100 according to Embodiment 2 of the present disclosure. Since the processing from steps S2301 to S2306 included in this flowchart is the same as the processing from steps S1301 to S1306 in FIG. 9, the description thereof is omitted. Also, the processing from steps S2310 to S2312 is the same as the processing from steps S1307 to S1309 in FIG. 9, so description thereof will be omitted.
  • Step S2307) The second information acquisition unit 106 acquires the switching signal from the second input unit 109 according to the user's input operation. Then, the second information acquisition unit 106 outputs to the image generation unit 107 the content corresponding to the acquired switching signal among the second information of each of the one or more known materials shown in the known material region d2.
  • the switching signal is a signal indicating the radio button d21 and the map display button d22 selected by the user.
  • the content corresponding to the switching signal is, for example, the compositional formula, firing temperature, or literature information among the second information of each of the one or more known materials.
  • Step S2308 The image generation unit 107 generates a map indicating the content of the second information of each of the one or more known materials acquired from the second information acquisition unit 106 and corresponding to the switching signal.
  • Step S2309 The image generation unit 107 displays the map generated in step S2309 on the display unit 108.
  • the second information selected in step S2310 may be the second information shown in the known material area d2, or may be the second information corresponding to notations such as dots shown on the map.
  • Embodiment 3 An image displayed by the generating process display device 100 according to the third embodiment will be described.
  • the production process display device 100 according to Embodiment 3 updates the production probability of each of the plurality of types of baking methods based on the holding device.
  • 12A and 12B are diagrams showing an example of display on the generating process display device 100 according to the third embodiment.
  • the image generation unit 107 displays the process-related image P1 of FIG. 8C according to Embodiment 1 including method selection buttons d41 corresponding to each of the five types of baking methods.
  • method selection buttons d41 corresponding to five types of firing methods, Ball-Milling, Solid-State, Liquid-State, Flux, and Arc-Melting.
  • These method selection buttons d41 indicate to the production process display device 100 that the baking method that can be used by the baking apparatus owned by the user of the production process display device 100 is the baking method corresponding to the method selection button d41. It is a button for notification.
  • a baking device owned by a user in this way is also called a owned device.
  • the user selects one or more method selection buttons d41 by performing an input operation on the second input unit 109.
  • FIG. the second information acquisition unit 106 acquires a signal corresponding to the input operation from the second input unit 109, and specifies one or more selected method selection buttons d41 based on the signal.
  • the second information acquisition unit 106 identifies the baking method corresponding to each of the selected one or more method selection buttons d41, and notifies the image generation unit 107 of it.
  • the image generation unit 107 updates the generation probability of each of the five baking methods according to the notified one or more baking methods.
  • the image generation unit 107 updates each generation probability in the generation probability area d3 of the process-related image P1. . That is, the image generation unit 107 updates the generation probabilities of Ball-Milling and Liquid-State, and deletes the generation probabilities of other firing methods, Solid-State, Flux, and Arc-Melting. For example, the image generation unit 107 multiplies the generation probabilities of Solid-State, Flux, and Arc-Melting by the weight "0" to update the generation probabilities to "0". Delete from area d3.
  • the image generation unit 107 updates the production probabilities of Ball-Milling and Liquid-State so that the sum of the production probabilities becomes 100 while maintaining the ratio between the production probabilities. .
  • the image generator 107 updates the generation probability a by a ⁇ 100/(a+b) ⁇ . That is, the image generation unit 107 updates the generation probability a by multiplying the generation probability a by the weight “100/(a+b)”.
  • the image generator 107 updates the generation probability b by b ⁇ 100/(a+b) ⁇ .
  • the image generation unit 107 updates the generation probability b by multiplying the generation probability b by the weight “100/(a+b)”.
  • the generation probabilities of Ball-Milling and Liquid-State updated in this way are displayed in the generation probability area d3 in the process-related image P1.
  • the image generator 107 may update the contents shown in the known material area d2 when one or more method selection buttons d41 are selected. For example, as described above, the Ball-Milling and Liquid-State method selection buttons d41 are selected. In this case, the image generation unit 107 leaves only the second information about each of Ball-Milling and Liquid-State out of the second information of each of the one or more known materials shown in the known material region d2, and leaves the remaining The second information may be deleted.
  • FIG. 13 is a flowchart showing an example of the overall flow of processing of the generating process display device 100 according to Embodiment 3 of the present disclosure. Since the processing from steps S3301 to S3306 included in this flowchart is the same as the processing from steps S1301 to S1306 in FIG. 9, description thereof will be omitted. Further, the processing from steps S3311 to S3313 is also the same as the processing from steps S1307 to S1309 in FIG. 9, so description thereof will be omitted.
  • Step S3307 The second information acquisition unit 106 identifies one or more firing methods that can be used with the device, based on the signal acquired from the second input unit 109 . Further, the second information acquisition unit 106 notifies the image generation unit 107 of one or more firing methods that can be used with the device.
  • Step S3308 The image generation unit 107 updates the generation probabilities of each of the five types of baking methods based on one or more types of baking methods that can be used by the device, notified by the second information acquisition unit 106 .
  • Step S3309) The image generator 107 reconstructs the process-related image P1 based on the updated generation probability.
  • This reconstructed process-related image P1 shows the updated production probabilities of one or more firing methods that can be used in the possessed apparatus, and does not show the production probabilities of other firing methods.
  • the contents shown in the known material area d2 may also be updated.
  • the image generation unit 107 leaves only the second information indicating one or more firing methods available in the holding device among the second information of each of the one or more known materials shown in the known material area d2, The remaining second information may be deleted.
  • the second information indicates, for example, the compositional formula, crystal structure, firing temperature, firing method, characteristic values, literature information, usage, and the like.
  • Step S3310 The image generation unit 107 displays the reconstructed process-related image P1 on the display unit 108.
  • the image generation unit 107 in the present embodiment identifies the types of baking methods that can be used in the baking apparatus owned by the user, and determines the generation probabilities of each of the one or more types of baking methods. By weighting each of the production probabilities of one type of baking method and the remaining other production probabilities differently, the production probabilities of each of the one or more types of baking methods shown by the process-related image P1 are obtained. update. As a result, it is possible to increase the probability of creating a firing method that can be used with the firing device owned by the user, and decrease the probability of creating a firing method that can only be performed with a firing device that the user does not have.
  • a weight greater than 1 is assigned to the production probability of the baking method that can be used by the baking apparatus owned by the user, and the production probability of the other baking methods is given by A weight of 0 is assigned.
  • the production probabilities of the remaining other firing methods may be weighted greater than zero.
  • FIG. 14 is a diagram showing another example of the process-related image P1 after the method selection button d41 in FIG. 12A is selected.
  • the Ball-Milling and Liquid-State method selection buttons d41 are selected.
  • the image generation unit 107 replaces the known material region d2 of the process-related image P1 with the quotation request region d5 without updating the generation probability of each of the five types of baking methods.
  • a quote request button d33 is displayed. This estimate request button d33 is associated with a firing method in the table d31 that is not possible with the owned apparatus.
  • the second input unit 109 transmits a signal indicating the selected estimate request button d33 to the second information acquisition unit. 106.
  • the second information acquisition unit 106 identifies the firing method associated with the selected estimate request button d33 based on the signal, and determines the firing method related to the estimate request for the firing apparatus capable of the firing method. Get RFQ data.
  • the quotation request data may be stored in the first storage unit 301 , the second storage unit 302 , the internal memory of the generation process display device 100 , or a server or storage connected to the network 401 .
  • Second information acquisition unit 106 then outputs the quotation request data to image generation unit 107 .
  • the image generation unit 107 When the image generation unit 107 acquires the quotation request data from the second information acquisition unit 106, the known material region d2 in the process-related image P1 is replaced with the quotation request region d5 indicating the quotation request data. Then, the image generation unit 107 displays the process-related image P1 including the quotation request area d5 on the display unit 108.
  • FIG. 1 When the image generation unit 107 acquires the quotation request data from the second information acquisition unit 106, the known material region d2 in the process-related image P1 is replaced with the quotation request region d5 indicating the quotation request data. Then, the image generation unit 107 displays the process-related image P1 including the quotation request area d5 on the display unit 108.
  • the generation probability for each type of firing method is derived and displayed, and the firing method corresponding to the method selection button d41 that was not selected is An estimate request button d33 is displayed.
  • the estimate request button d33 is selected, the estimate request data for requesting an estimate of the baking apparatus capable of the firing method associated with the selected estimate request button d33 is displayed in the estimate request area d5, which is the estimate request window. to be displayed.
  • the second information acquiring unit 106 in the present embodiment can perform firing by the remaining firing methods, excluding the firing method specified in step S3307, among the one or more firing methods.
  • the image generation unit 107 displays an image showing the estimate information on the display unit 108 .
  • the quotation information is, for example, the quotation request data described above, and the image showing the quotation information is the quotation request area d5 in the process-related image P1. Accordingly, the user can easily request an estimate for a baking apparatus that the user does not own, according to the displayed estimate information.
  • FIG. 4 An image displayed by the generating process display device 100 according to the fourth embodiment will be described.
  • the production process display device 100 according to Embodiment 4 updates the contents of the known material area d2 based on the estimated required time.
  • 15A and 15B are diagrams showing an example of display on the generating process display device 100 according to the fourth embodiment.
  • the image generation unit 107 displays the process-related image P1 of FIG. 8C according to Embodiment 1 including the required time prediction button d42.
  • the image generation unit 107 determines that the estimated required time d23 for each of the one or more known materials is included in the known material region d2 as shown in FIG. 15B. update its known material region d2 so that The estimated required time d23 is, for example, the time estimated to be required for baking, synthesizing, or producing the material.
  • the second information acquisition unit 106 obtains the first combined information 2a and the second combined information 2b of each of the one or more known materials related to the target material. 2 Get information. This second information indicates the composition formula other than the estimated required time, the firing method, the literature information, the application, and the like. Then, the second information acquisition unit 106 outputs the second information acquired for each of the one or more known materials to the image generation unit 107 .
  • the image generation unit 107 generates a process-related image P1 including a required time prediction button d42 and a known material region d2 indicating second information acquired for each of one or more known materials, and generates the process-related image P1. P1 is displayed on the display unit 108 .
  • the user selects the required time prediction button d42 by performing an input operation on the second input unit 109.
  • the second information acquisition unit 106 acquires a signal corresponding to the input operation from the second input unit 109, and based on the signal, determines that the required time prediction button d42 has been selected.
  • the second information acquisition unit 106 acquires information indicating the estimated required time from the second combined information 2b of each of the one or more known materials related to the target material, and outputs the information to the image generation unit 107 .
  • the image generating unit 107 acquires the information indicating the estimated required time for each of the one or more known materials from the second information acquiring unit 106
  • the image generating unit 107 updates the known material region d2 of the process-related image P1 as shown in FIG. 15B. . That is, as shown in FIG. 15A, the known material region d2 indicating attributes such as the composition formula and crystal structure of each of one or more known materials for each type of firing method is updated to the known material region d2 shown in FIG. 15B. be done.
  • the estimated required time for each of one or more known materials is shown as part of the known material attributes. Further, in the known material region d2 shown in FIG.
  • composition formulas and attributes of each of the one or more known materials are not classified by type of firing method, but are arranged in ascending order of the estimated required time of the known materials. are shown.
  • the attributes of the known material include the estimated time required for the known material, the firing method, literature information, usage, and the like. That is, the composition formulas and attributes of the known materials are arranged in descending order of the estimated required time. Note that when there are a large number of sets including the composition formula and attributes of the known material shown in the known material region d2, each set of one or more known materials for each type of firing method has a short estimated required time. They may be arranged and displayed in order.
  • the second information acquiring unit 106 acquires information indicating the estimated required time.
  • Information indicating the estimated required time may be acquired.
  • the second information acquisition unit 106 acquires information indicating the estimated required time before the required time prediction button d42 is selected, that is, when the process-related image P1 shown in FIG. 15A is displayed. , the information is output to the image generation unit 107 .
  • the image generation unit 107 updates the known material region d2 as shown in FIG. 15B based on information indicating the estimated required time acquired in advance.
  • the image generation unit 107 displays the process-related image P1 including the updated known material region d2 on the display unit 108.
  • FIG. 16 is a flowchart showing an example of the overall flow of processing of the generating process display device 100 according to Embodiment 4 of the present disclosure. Since the processing from steps S4301 to S4306 included in this flowchart is the same as the processing from steps S1301 to S1306 in FIG. 9, the description thereof is omitted. Further, the processing from steps S4310 to S4312 is also the same as the processing from steps S1307 to S1309 in FIG. 9, so description thereof will be omitted.
  • Step S4307) Based on the signal obtained from the second input unit 109, the second information obtaining unit 106 determines that the required time prediction button d42 has been selected. That is, the second information acquisition unit 106 accepts an instruction to predict required time. As a result, the second information acquiring unit 106 acquires the information indicating the estimated required time associated with each composition formula of one or more known materials related to the target material as the second combined information in the second storage unit 302. 2b and output to the image generation unit 107 .
  • Step S4308) The image generation unit 107 updates the known material region d2 of the process-related image P1 based on the information indicating the estimated required time acquired from the second information acquisition unit 106.
  • FIG. By this update, the known material area d2 shows the estimated required time for each of the one or more known materials in association with the composition formula of the known material.
  • Step S4309) The image generation unit 107 displays the process-related image P1 including the updated known material region d2 on the display unit 108.
  • the attribute indicated by the second information of each of the one or more known materials includes the estimated required time, which is the time required for firing the known material.
  • the image generation unit 107 arranges the second information of each of the one or more known materials in the order according to the estimated required time indicated by the second information, and arranges the one or more second information arranged in the order.
  • a process-related image P1 showing information is generated. This displays the estimated required times for each of the one or more known materials, allowing the user to estimate the time required to fire the target material based on those estimated required times.
  • the estimated required times are displayed in order, the user can easily grasp the maximum value, the minimum value, or the variance of the estimated required time, and the Time can be predicted better.
  • each attribute of one or more known materials includes a firing method
  • the user can easily determine which firing method can reduce the time required to fire the target material. can do. For example, even if the user intends to use the baking method with the highest production probability shown in the process-related image P1 for baking the target material, if the estimated required time for that baking method is long, the next highest production probability It can be determined that the firing method is used to fire the target material. In other words, the user can select the firing method of the target material by considering the time required for firing.
  • FIG. 5 An image displayed by the generating process display device 100 according to the fifth embodiment will be described.
  • the generation process display device 100 according to Embodiment 5 generates a process-related image P1 including a map instead of the known material region d2. Then, when a dot corresponding to the compositional formula of a known material on the map is selected, the production process display device 100 displays the reference image P2 corresponding to the compositional formula on the process-related image P1.
  • 17A, 17B, and 17C are diagrams showing an example of display on the generating process display device 100 according to Embodiment 5.
  • FIG. 5 An image displayed by the generating process display device 100 according to the fifth embodiment will be described.
  • the generation process display device 100 according to Embodiment 5 generates a process-related image P1 including a map instead of the known material region d2. Then, when a dot corresponding to the compositional formula of a known material on the map is selected, the production process display device 100 displays the reference image P2 corresponding to the composition
  • the image generator 107 generates an initial image P0 including a plurality of map selection buttons d43 for selecting the map type, as shown in FIG. 17A.
  • the initial image P0 includes a target composition area d1 that is empty for receiving input of a target composition formula, a map selection button d43 for selecting a composition formula information map, and a sintering temperature map selection button d43. map selection button d43.
  • the user inputs the target composition formula in the target composition area d1 by performing an input operation on the first input unit 101, and further selects one of the two map selection buttons d43.
  • the first information acquisition unit 102 acquires a signal indicating the target composition formula according to the input operation from the first input unit 101 as first information, and selects the map selection button d43 according to the input operation. Get a map selection signal that indicates the type of map supported.
  • the map selection signal indicates a composition formula information map as the type of map.
  • the image generation unit 107 acquires the first information and the map selection signal indicating the composition formula information map from the first information acquisition unit 102, generates the process-related image P1 shown in FIG. 108.
  • the process-related image P1 shown in FIG. 17B includes a composition formula information map m11 instead of the known material region d2.
  • the vertical axis of the composition formula information map m11 indicates the composition difference
  • the horizontal axis of the composition formula information map m11 indicates the composition formula.
  • the composition difference is the difference between the composition descriptor corresponding to the target composition formula and the composition descriptor corresponding to the composition formula of the known material, and is the distance between the vectors.
  • composition formula information map m11 the composition formula of each of one or more known materials related to the target material is arranged according to the analysis result of the principal component analysis.
  • dots corresponding to each of one or more known materials related to the target material are arranged at positions corresponding to the compositional difference and compositional formula of the known material.
  • the image generation unit 107 includes the firing temperature map m12 instead of the known material region d2 in the process-related image P1, as shown in FIG. 17C. 108.
  • the vertical axis of the firing temperature map m12 indicates the firing temperature
  • the horizontal axis of the firing temperature map m12 indicates the composition difference.
  • dots corresponding to each of one or more known materials related to the target material are arranged at positions corresponding to firing temperatures and composition differences of the known materials.
  • the user selects a dot on the composition formula information map m11 or the firing temperature map m12 by performing an input operation on the second input unit 109.
  • the second information acquisition unit 106 acquires a signal corresponding to the input operation from the second input unit 109 and identifies the selected dot based on the signal. Further, the second information acquiring unit 106 acquires from the first database a paragraph describing a method of sintering the compositional formula of the known material corresponding to the specified dot. Second information acquisition section 106 then outputs the paragraph to image generation section 107 . After obtaining the paragraph from the second information obtaining unit 106, the image generation unit 107 generates a reference image P2 indicating the paragraph and includes it in the process-related image P1. The image generation unit 107 displays the process-related image P1 including the reference image P2 on the display unit 108.
  • FIG. 18 is a flowchart showing an example of the overall flow of processing of the generating process display device 100 according to Embodiment 5 of the present disclosure. Since the processing from steps S5302 to S5304 included in this flowchart is the same as the processing from steps S1302 to S1304 in FIG. 9, the description thereof is omitted. Further, the processing from steps S5306 to S5309 is also the same as the processing from steps S1306 to S1309 in FIG. 9, so description thereof will be omitted.
  • Step S5301 The first input unit 101 outputs first information, which is a signal indicating the target composition formula to be displayed, and a map selection signal indicating the type of map, to the first information acquisition unit 102 by the user's input operation. .
  • First information acquisition section 102 outputs the first information and the map selection signal to image generation section 107 .
  • Step S5305) The image generation unit 107 acquires the second information of each of the one or more known materials related to the target material acquired from the second information acquisition unit 106, and the map selection signal acquired from the first information acquisition unit 102. Generate a map. In other words, the image generation unit 107 generates a map by arranging the second information in a display form according to the type of map indicated by the map selection signal.
  • the map is, for example, the composition formula information map m11 or the firing temperature map m12 described above.
  • the image generator 107 then generates a process-related image P1 including the map.
  • step S5307 corresponds to the dots corresponding to the compositional formulas of the known materials on the map described above.
  • the image generation unit 107 uses the composition formula information map m11 representing the relationship between the composition formulas indicated by the second information of each of one or more known materials, or , a firing temperature map m12 showing the relationship between the compositional formula and the attribute indicated by the second information of each of the one or more known materials is generated and displayed on the display unit .
  • a large amount of second information is displayed as a map instead of being displayed as a list, so that the user can easily find the desired second information from the second information.
  • the user can easily select the desired second information and easily view the reference image P2 corresponding to the second information.
  • the generation process display device has been described above based on Embodiments 1 to 5, the present disclosure is not limited to those embodiments. Various modifications conceived by those skilled in the art may be included in the present disclosure as long as they do not deviate from the spirit of the present disclosure.
  • the present disclosure may also include a configuration constructed by combining constituent elements of a plurality of mutually different embodiments.
  • the generating process display device 100 displays the process-related image P1 and the reference image P2 on the display unit 108, but outputs information contained in those images without displaying them. good too.
  • the productive process display device 100 may be called a productive process output device.
  • the production process output device is a device that outputs information about the production process of a material, and includes a first information acquisition unit 102 that acquires first information indicating the composition formula of the target material, and a composition indicated by the first information.
  • a second information acquisition unit 106 that acquires second information indicating the compositional formula and attributes of the known material from a material database, and one or more firing methods.
  • a probability derivation unit 105 that derives the probability that the target material having the composition formula indicated by the first information will be fired by the firing method of the type as the generation probability based on the database, and one or more of known materials
  • An output unit for outputting each of the second information and each of the generation probabilities of the one or more firing methods as information on the generation process.
  • the second information acquisition unit 106 fires the known material having the composition formula indicated by the selected second information.
  • the second information acquisition unit 106 of the generation process output device does not have to acquire some areas in the above-mentioned document from the database. Furthermore, the output unit does not have to output the information indicating the description content of some areas in the document. Even in such a case, except for the information shown in the reference image P2, the same effect as the above-described production process display device 100 can be obtained with respect to the information regarding the production process.
  • the probability derivation unit 105 derives the production probability of each of the five types of firing methods for the target material by inputting the composition descriptor of the target composition formula into the firing predictor.
  • the probability derivation unit 105 uses one firing predictor, but may use a plurality of firing predictors.
  • each of the five firing predictors outputs the generation probability of one type of firing method previously associated with that firing predictor.
  • the probability derivation unit 105 derives the production probability of each of the five types of firing methods for the target material by inputting the composition descriptor of the target composition formula into each of the five firing predictors.
  • the firing predictor calculates the generation probability of each of the five types of firing methods for the input of the composition descriptor of the target composition formula consisting of 72 or less or 86 or less element species. Output.
  • the firing predictor may output the generation probability for the input of the composition descriptor of the composition formula consisting of two kinds of element species.
  • the probability derivation unit 105 inputs a composition descriptor corresponding to each combination of two element species included in the target composition formula into the firing predictor, thereby determining five types of firing methods for that combination.
  • the probability derivation unit 105 calculates the generation probability of the firing method for the target composition formula from the generation probability of the firing method derived for each combination for each of the five types of firing methods. For example, in the case of the target composition formula “BiCuSO”, the combinations of two element species included in the target composition formula are “BiCu”, “BiS”, “BiO”, “CuS”, “CuO”, and "SO”.
  • the probability derivation unit 105 inputs the composition descriptor corresponding to the combination into the firing predictor for each combination, thereby deriving the generation probability of each of the five types of firing methods for that combination.
  • the probability deriving unit 105 calculates the generation probability of the predetermined firing method for “BiCu”, the generation probability of the predetermined firing method for “BiS”, the generation probability of the predetermined firing method for “BiO”, and the generation probability of the predetermined firing method for “BiO”.
  • a predetermined firing method for the target composition formula “BiCuSO” using the generation probability of a predetermined firing method for “CuO” and the generation probability of a predetermined firing method for “SO” Calculate the generation probability of
  • the average of the production probabilities derived for each combination may be used to calculate the production probability of the predetermined firing method for the target composition formula, or a weighted addition of these production probabilities may be used.
  • the generation probability of each of the five types of firing methods is derived and displayed, but the types of firing methods are not limited to five types, and may be one or more and less than five types. Well, it may be six or more types. Furthermore, the type of firing method is not limited to Ball-Milling, Solid-State, Liquid-State, Flux, and Arc-Melting, and may be other types. Furthermore, the names of the firing methods Ball-Milling, Solid-State, Liquid-State, Flux, and Arc-Melting are only examples, and the names are not limited to these names. For example, Ball-Milling may be Ball-Mill.
  • the first database in each of the above embodiments has a plurality of article data 1 as documents, but the documents are not limited to article data 1 indicating articles, but data indicating magazines, books, experimental reports, etc. may be
  • the first combined information 2a and the second combined information 2b are independent of each other, but they may be configured integrally.
  • the attribute indicated by the second combined information 2b in each of the above embodiments includes the location of the paragraph in the article data 1, but may include the paragraph itself.
  • the second information acquisition unit 106 can acquire paragraphs from the second combined information 2b without using the article data 1 of the first database.
  • the second information for each of a plurality of known materials is displayed, but the number of pieces of second information may be one.
  • the generating process display device 100 of the first to fifth embodiments can be said to be the following generating process display device or generating process output device.
  • the production process display device is a production process display device that displays information about the production process of a material, and includes a first information acquisition unit that acquires first information indicating a compositional formula of a target material, and a first information that is indicated by the first information.
  • a second information acquisition unit that acquires second information indicating the composition formula and attributes of each of the one or more known materials related to the composition formula from a material database, and one or more firing methods, respectively.
  • a probability derivation unit that derives the probability that the target material having the composition formula indicated by the first information will be fired by the type of firing method as a generation probability based on the database; and the one or more known an image generation unit that generates a first image showing the second information of each material and the generation probability of each of the one or more firing methods and displays the first image on a display unit, and acquires the second information.
  • the part further comprises a known material having a composition formula indicated by the selected second information when one of the second information of each of the one or more known materials indicated by the first image is selected.
  • a partial area in a document describing a material firing method is acquired from the database, and the image generation unit further generates an image showing the description content of the partial area in the document as a second image. and displayed on the display unit.
  • the first image is also called a process-related image
  • the second image is also called a reference image.
  • the knowledge and experience of the user can be used to select a firing method and set experimental conditions with higher accuracy. For example, if the production probability of each of multiple types of firing methods is shown in the first image (i.e., the process-related image), the user can select the type of firing method corresponding to the highest generation probability to produce the desired material. can be determined to be fired. Furthermore, for example, if the user does not have a firing apparatus capable of the type of firing method corresponding to the highest generation probability, the user confirms the generation probability corresponding to the firing method possible with the firing apparatus that is owned. be able to.
  • the user can make a decision that even if the firing method corresponds to the lowest generation probability, if the generation probability is, for example, 30% or higher, the firing of the target material by that firing method will be attempted. .
  • the composition formula and attributes of each of the one or more known materials related to the target material for example, the composition formula and attributes of each of the one or more known materials having a composition formula similar to the target material are displayed in the first image can be shown.
  • the attribute may be a matter of firing, synthesizing or producing a known material.
  • Firing of materials is a subordinate concept of synthesis of materials, and synthesis of materials is a subordinate concept of generation of materials. As such, calcination may be referred to as synthesis or generation.
  • the probability derivation unit may derive the generation probability based on the combination and composition ratio of element species included in the composition formula indicated by the first information.
  • the database may indicate, for each of a plurality of materials, the composition formula, firing method, and attributes of the material in association with each other.
  • the probability deriving unit uses the database so as to output the generation probability of each of the one or more firing methods for the material having the composition formula in response to the input of the descriptor indicating the composition formula.
  • the generation probability of each of the one or more firing methods for the target material may be derived by inputting a descriptor indicating the compositional formula of the target material into the firing predictor trained by the method.
  • the bake predictor is a neural network or the like.
  • the generation probability of each of N types of firing methods (N is an integer of 3 or more) is derived by the probability derivation unit, and among the N types of firing methods, the probability is higher than the remaining other firing methods.
  • the image generation unit performs the N types of baking.
  • the first image may be generated showing the generation probability of each of the M types of firing methods and not showing the generation probabilities of the remaining other firing methods.
  • the M types of firing methods may be one type of firing method.
  • the generation probability of one type of firing method among a plurality of firing methods is a threshold value of 80% or more, only the generation probability of that one type of firing method is displayed, and the remaining firing methods are displayed. Generation probabilities are not displayed. Therefore, the user can easily find the most likely firing method for the target material.
  • the generation probabilities of each of the three or more types of baking methods for example, if the sum of the top two generation probabilities is a threshold value of 90% or more, only the top two generation probabilities are displayed, and the remaining other generation probabilities are displayed. Probability is not shown. Note that the top two generation probabilities are two generation probabilities that are higher than any other generation probabilities. Therefore, it is possible to omit from the first image the production probability that the target material is unlikely to be fired, and to suppress unnecessary information provision to the user.
  • Each of the one or more firing methods may be Ball-Milling, Solid-State, Liquid-State, Flux, or Arc-Melting.
  • Ball-Milling Solid-State, Liquid-State, Flux, and Arc-Melting are only examples, and the names are not limited to these names.
  • Ball-Milling may be Ball-Mill.
  • a part of the document may be a sentence, a paragraph, or a paragraph containing the content of the firing method.
  • the image generation unit emphasizes a word representing the baking method indicated as the attribute by the second information selected from the partial area in the document more than other remaining words.
  • the second image may be generated.
  • the word representing the firing method is highlighted, so the user can easily find the word representing the firing method from a part of the document. Furthermore, the user can search around the word to easily find the description related to the firing method.
  • the image generation unit may generate, for each type of firing method, the first image indicating the second information of each of the one or more known materials generated by the type of firing method. .
  • the user can easily find out from the first image the second information of each of the one or more known materials fired by the firing method with the highest production probability for the target material, for example. That is, even if a large amount of second information is displayed, the user can easily find useful second information regarding firing of the target material.
  • the image generation unit further includes a map representing the relationship of composition formulas indicated by the second information of each of the one or more known materials, or the second information of each of the one or more known materials.
  • a map showing the relationship between the indicated compositional formula and the attribute may be generated and displayed on the display unit.
  • the second information is displayed on the map, so that the user can easily find the desired second information from the second information. be able to.
  • the user can easily select the desired second information and easily view the second image (that is, the reference image) corresponding to the second information.
  • the attributes of the known material indicated by the second information include a crystal structure possessed by the known material, a firing temperature for firing the known material, a characteristic value indicating the degree of characteristics of the known material, and the known material. and at least one of the literature information for identifying the literature in which the is described and the use of the known material.
  • the user can know the attributes of known materials, and can easily guess the firing conditions of the target material based on those attributes.
  • the image generation unit arranges the second information of each of the one or more known materials in an order according to the composition formula or the attribute indicated by the second information, and The first image may be generated to represent one or more of the second information.
  • the plurality of pieces of second information are arranged in the order of the composition formula, which is the order corresponding to the composition formula indicated by the second information, or in the order of the attribute, which is the order corresponding to the attribute indicated by the second information. arrayed.
  • the second information indicating the compositional formula of a known material that is close to or similar to the compositional formula of the target material is arranged first.
  • the attribute is the firing temperature of a known material
  • the second information indicating the higher firing temperature is arranged first in the order of the attribute. This allows the user to easily find the desired second information from the second information.
  • the image generation unit further specifies the types of baking methods that can be used by the baking apparatus owned by the user, and selects the specified type among the generation probabilities for each of the one or more types of baking methods. and the remaining other production probabilities are weighted differently from each other, the production probabilities for each of the one or more firing methods shown by the first image. may be updated.
  • the second information acquisition unit further relates to an estimate of a firing apparatus capable of firing by the remaining firing method excluding the specified type of firing method among the one or more firing methods.
  • the image generation unit may further display an image showing the estimate information on the display unit.
  • the user can easily request an estimate for a baking apparatus that the user does not own, according to the displayed estimate information.
  • the attribute indicated by the second information of each of the one or more known materials includes an estimated required time that is the time required to bake the known material
  • the image generation unit generates the one or more known materials arranging each of the second information in an order according to the estimated required time indicated by the second information, and generating the first image showing one or more of the second information arranged in the order good too.
  • the estimated required time for each of one or more known materials is displayed, so the user can predict the time required for firing the target material based on the estimated required times.
  • the user can easily grasp the maximum value, the minimum value, or the variance of the estimated required time, and the Time can be predicted better.
  • each attribute of one or more known materials includes a firing method, the user can easily determine which firing method can reduce the time required to fire the target material. can do. For example, even if the user intends to use the baking method with the highest production probability shown in the first image to bake the target material, if the estimated required time for that baking method is long, the baking method with the next highest production probability will be used. It can be determined that the method is used for firing the target material. In other words, the user can select the firing method of the target material by considering the time required for firing.
  • the production process output device is a production process output device that outputs information about the production process of the material, and is a first information acquisition unit that acquires first information indicating the composition formula of the target material, and the first information
  • a second information acquisition unit that acquires, from a material database, second information indicating the composition formula and attributes of each of the one or more known materials associated with the indicated composition formula, and one or more firing methods.
  • a probability derivation unit for deriving, based on the database, the possibility that the target material having the composition formula indicated by the first information will be fired by the firing method of the type, as a generation probability, for each of and an output unit that outputs the second information for each of the known materials and the generation probability for each of the one or more firing methods as information related to the generation process, wherein the second information acquisition unit is Furthermore, when one of the second information of each of the one or more known materials output is selected, a method of firing a known material having a composition formula indicated by the selected second information is described. A part of the document containing the reference is obtained from the database, and the output unit further outputs information indicating the description content of the part of the document.
  • the production process output device is a production process output device that outputs information about the production process of the material, and includes a first information acquisition unit that acquires first information indicating the composition formula of the target material, and the first information A second information acquisition unit that acquires second information indicating the composition formula and attributes of each of the one or more known materials associated with the indicated composition formula from a material database, and one or more firing methods.
  • a probability derivation unit for deriving, based on the database, the possibility that the target material having the composition formula indicated by the first information will be fired by the firing method of the type, as a generation probability, for each of and an output unit that outputs the second information for each of the known materials and the generation probability for each of the one or more firing methods as information on the generation process.
  • production parameters are derived that indicate the degree to which each of multiple types of production processes is required to produce a target material using multiple raw materials.
  • the target material is also called a composite material or a mixed material.
  • a single raw material may be used to produce the target material instead of multiple raw materials.
  • the production probability of the firing method is derived.
  • the production process is not limited to the firing method.
  • the generation probability is used as an example of the generation parameter as in the first to fifth embodiments, but the generation parameter may be the order shown in FIG. 8A, for example. Such ranking indicates the relative degree to which each of the multiple production processes is required.
  • the first information in Embodiments 1 to 5 above indicates the composition formula of the target material
  • the second information is the composition formula of the known material related to the target material and the composition formula is linked to the composition formula. attribute.
  • the first information in this embodiment indicates the name of one or more raw materials for producing the target material.
  • the name is information for identifying the raw material, and may be, for example, the compositional formula of the raw material. It can also be said that the first information indicates one or more raw materials by the name of one or more raw materials.
  • the second information in the present embodiment is the name of one or more raw materials (hereinafter also referred to as related raw materials) for producing a known material related to one or more raw materials of the target material, and the one or more associated raw materials and linked attributes. It can also be said that the second information indicates one or more related raw materials by the name of one or more related raw materials.
  • Embodiments 1 to 5 are denoted by the same reference numerals as in Embodiments 1 to 5, and detailed descriptions thereof are omitted. Further, among the terms used in the present embodiment, the same definitions and meanings as in Embodiments 1 to 5 are applied to the same terms as those in Embodiments 1 to 5 unless otherwise specified. be.
  • FIG. 19 is a diagram showing an example of the configuration of a generated process search system 10a according to this embodiment.
  • the generated process search system 10a includes a generated process display device 100a, a display unit 108, an input unit 110, a database construction device 200a, a first storage unit 301, and a second storage unit 302a.
  • the generation process display device 100a, database construction device 200a, first storage unit 301, and second storage unit 302a are connected to each other via a network 401 such as the Internet.
  • the production process display device 100a is configured as, for example, a computer such as a personal computer or a server, and displays information on the production process, which is the process for producing the target material, on the display unit 108.
  • the generating process display device 100a displays an image showing information about the generating process on the display unit 108, the information may be output without displaying the image.
  • the generated process display device 100a may be called a generated process output device.
  • the image is also called the first image or the process-related image.
  • the display unit 108 displays an image showing information about the above-described generation process according to a signal output from the generation process display device 100a.
  • the input unit 110 receives an input operation by the user of the generating process display device 100a, and outputs a signal corresponding to the input operation to the generating process display device 100a.
  • the display unit 108 and the input unit 110 are configured independently of each other, but they may be configured integrally like a touch panel.
  • the generating process display device 100a does not include the display unit 108 and the input unit 110, but may include them.
  • the database construction device 200a is configured as, for example, a computer such as a personal computer or a server, and constructs a second database used in the generation process display device 100a.
  • the database construction device 200a reads the first database stored in the first storage unit 301 from the first storage unit 301 via the network 401, and uses the first database to construct the second database. .
  • the database construction device 200a stores the constructed second database in the second storage unit 302a via the network 401.
  • the first storage unit 301 is a recording medium for storing the first database, as in the first to fifth embodiments.
  • the second storage unit 302a is a recording medium for storing the second database. Both the first database and the second database are databases related to materials. Details of the second database will be described later.
  • These recording media are, for example, hard disk drives, RAMs, ROMs, or semiconductor memories. Note that such a recording medium may be volatile or nonvolatile.
  • first storage unit 301 and the second storage unit 302a are arranged outside the database construction device 200a in the example shown in FIG. 19, they may be provided in the database construction device 200a. Also, the first storage unit 301 and the second storage unit 302a may be directly connected to the database construction device 200a without going through the network 401. FIG. Also, the first database and the second database are stored in different recording media, but may be stored in the same recording medium. Furthermore, the set of the first database and the second database may be treated as one database.
  • FIG. 20 is a block diagram showing an example of the functional configuration of the generating process display device 100a according to this embodiment.
  • the generated process display device 100a uses the first database stored in the first storage unit 301 and the second database stored in the second storage unit 302a.
  • the first database contains a plurality of article data 1 .
  • Each of the plurality of article data 1 is associated with the material name of the main material (that is, known material) described in the article data 1 .
  • the material name may be a composition formula of a known material.
  • the material name may be the name or formula of one or more raw materials to produce the known material.
  • the second database includes a plurality of first generated information 2c and a plurality of second generated information 2d.
  • Each of the plurality of first generation information 2c is the name of one or more raw materials for generating the known material of the material name associated with the article data 1, and the known material is generated from the one or more raw materials.
  • the name of the generation process for The first generation information 2c may indicate the names of multiple generation processes.
  • Each of the plurality of second generation information 2d is the name of one or more raw materials for generating the known material of the material name associated with the article data 1, and the name of the known material indicated in the article data 1 Attributes are shown in association with each other.
  • the attribute of the known material is the document information of the article data 1 that shows the production process for producing the known material from the one or more raw materials, or describes the production of the known material in the article data 1 It may be the location of a paragraph or the like.
  • the attributes of the known material may be the crystal structure, property values, application, process conditions, etc. of the known material.
  • the process conditions are conditions used in the production process of the known material, such as temperature and time.
  • the attributes of the known material may include all or part of each of the above items.
  • the production process display device 100a is a device for displaying information about the production process of materials. Displays the production probability for each type of production process for materials.
  • the generation probability for each type of generation process for the target material is the degree to which the type of generation process is required to generate the target material for each of a plurality of types of predetermined generation processes. For example, the generation probability is expressed as a percentage. The higher the production probability of the production process for the target material, the higher the production probability indicates that the production process is required to produce the target material.
  • the generation process display device 100 a is connected to the first input section 101 and the second input section 109 included in the input section 110 and the display section 108 .
  • the generation process display device 100a includes a first information acquisition unit 102, a descriptor generation unit 103a, a generation learning unit 104a, a probability derivation unit 105, a second information acquisition unit 106, and an image generation unit 107. Prepare.
  • the generating process display device 100a does not include the first input section 101, the second input section 109, and the display section 108. good.
  • the first input unit 101 is a functional component included in the input unit 110, and receives a user's input operation, for example, outputs a signal indicating the name of one or more raw materials according to the input operation to the first input unit 101. Output to the information acquisition unit 102 .
  • First information acquisition section 102 receives a signal from first input section 101 . That is, the first information acquisition unit 102 acquires first information, which is a signal indicating the name of one or more raw materials. The first information acquisition unit 102 outputs the first information to the descriptor generation unit 103 a and the image generation unit 107 .
  • the descriptor generation unit 103a acquires the first information from the first information acquisition unit 102 and generates descriptors corresponding to the names of one or more raw materials indicated by the first information.
  • a descriptor is, for example, a vector that uniquely indicates a raw material. Also, the descriptor may be the same as the descriptor generated by the composition descriptor generator 103 of Embodiments 1-5.
  • the descriptor generation unit 103a may generate a descriptor of the organic molecular material as follows. That is, the descriptor generation unit 103a acquires the first information from the first information acquisition unit 102 and creates a descriptor corresponding to the organic molecular material indicated by the first information.
  • a descriptor is, for example, a vector that uniquely indicates an organic molecular material.
  • the descriptor generation unit 103a uses, as descriptors, parameters indicating the number of heteroatoms, the number of amide bonds, and the like in the organic molecular material, that is, parameters calculated from constituent elements, structures, and the like.
  • the descriptor generation unit 103a first describes the character string of the organic molecular material.
  • a description method called SMILES simple molecular input line entry system
  • PVA polyvinyl alcohol
  • the descriptor generation unit 103a digitizes this character string to generate a numerical vector configured by arranging numerical values such as the number of heteroatoms and the number of amide bonds of the organic molecular material. This numeric vector is the descriptor used as input for machine learning.
  • the descriptor of the organic molecular material may be generated as a One Hot Vector. For example, when there are only two candidates for organic molecular materials, polyvinyl alcohol and cellulose, a vector [0, 1] is generated as a descriptor for polyvinyl alcohol and a vector [1, 0] is generated as a descriptor for cellulose. good too.
  • Descriptor generating section 103 a outputs the descriptor to probability deriving section 105 and second information acquiring section 106 .
  • the generation learning unit 104a refers to the second database of the second storage unit 302a and learns the relationship between one or more raw materials and the generation process, thereby constructing a generation predictor that indicates the relationship.
  • the generation learning unit 104a uses a plurality of pieces of first generation information 2c included in the second database as teacher data.
  • Each of the plurality of pieces of first production information 2c indicates the name of one or more raw materials in association with the name of the production process for producing the known material from the one or more raw materials.
  • the generation learning unit 104a outputs the names of one or more raw materials indicated in the first generation information 2c to the descriptor generation unit 103a, and the descriptor generation unit 103a generates descriptors corresponding to the names of the one or more raw materials. 103a. Then, the generation learning unit 104a generates, for each of the plurality of first generated information 2c, one or more descriptors corresponding to one or more raw material names indicated in the first generated information 2c, and the first generated information 2c. Learning is performed using the name of the generation process shown in , as teacher data.
  • the generation learning unit 104a is a generation predictor that outputs the generation probability of each of a plurality of types of generation processes for generating ingredients from one or more raw materials in response to the input of one or more raw material descriptors. to build.
  • the generation probability of each of the plurality of types of generation processes indicates the degree to which that type of generation process is required to generate the target material using one or more raw materials corresponding to the one or more input descriptors. .
  • the multiple types of production processes are, for example, seven types of production processes, including stirring, mixing, sonication, freezing, melting, freeze-drying, impurity removal, and the like. Note that the multiple types of generation processes are not limited to these.
  • the generation learning unit 104a acquires one or more descriptors corresponding to one or more raw materials for generating the target material from the probability derivation unit 105
  • the generation learning unit 104a inputs the one or more descriptors to the generation predictor. obtains the respective production probabilities of the plurality of types of production processes for the target material from the production predictor.
  • Generation learning section 104 a then outputs these generation probabilities to probability derivation section 105 .
  • the generation learning unit 104a re-learns the generation predictor based on the difference between the post-update and the pre-update.
  • the probability derivation unit 105 acquires one or more descriptors corresponding to one or more raw materials for generating the target material from the descriptor generation unit 103a, and each of the plurality of types of generation processes is used to generate the target material. The required degree is derived as a generation probability. A generation predictor of the generation learning unit 104a is used to derive the generation probability. In other words, the probability derivation unit 105 regards each of the plurality of types of generation processes as the generation probability, which is the degree to which the type of generation process is required to generate the target material using one or more raw materials indicated in the first information. , based on the second database of the second storage unit 302a.
  • the probability derivation unit 105 inputs one or more descriptors acquired from the descriptor generation unit 103a to the generation predictor of the generation learning unit 104a. Then, the probability derivation unit 105 acquires the generation probability of each of the plurality of types of generation processes output from the generation predictor from the generation predictor. The probability derivation unit 105 outputs the generation probabilities of the multiple types of generation processes to the second information acquisition unit 106 and the image generation unit 107 .
  • the second information acquisition unit 106 acquires descriptors for each of one or more raw materials for generating the target material from the descriptor generation unit 103a. Then, the second information acquisition unit 106 acquires one or more pieces of second information necessary for image generation from the second database of the second storage unit 302a based on the acquired one or more descriptors. Further, second information acquisition section 106 outputs the second information to image generation section 107 . As a result, an image (that is, a process-related image) representing one or more pieces of second information is generated by the image generator 107 and displayed.
  • the second information acquired by the second information acquisition unit 106 includes the names of one or more related raw materials related to the one or more raw materials indicated by the first information, and the names of known materials generated from the one or more related raw materials. attribute. Its attributes include document information and the location of paragraphs.
  • the second information acquiring unit 106 obtains the first generated information 2c and the second generated information indicating one or more related raw materials among the plurality of first generated information 2c and the plurality of second generated information 2d included in the second database. By obtaining the information 2d, the second information is obtained. Note that the generation process indicated by the first generation information 2c of the acquired second information may be treated as an attribute of the known material.
  • the second information acquisition unit 106 retrieves the article corresponding to the selected second information.
  • a paragraph of data 1 may be obtained from the first database and output to the image generator 107 .
  • the image generation unit 107 acquires the first information from the first information acquisition unit 102 and generates an image showing the first information. In addition, the image generator 107 generates an image showing the generation probabilities acquired from the probability derivation unit 105 . In other words, the image generation unit 107 generates an image showing the generation probability of each of the seven types of generation processes for generating the target material from the one or more raw materials indicated in the first information. Also, the image generation unit 107 acquires one or more pieces of second information from the second information acquisition unit 106 and generates an image showing the second information. Each generated image described above is included in, for example, a process-related image and displayed on the display unit 108 .
  • the image generation unit 107 generates an image representing one or more pieces of second information and generation probabilities for each of a plurality of types of generation processes as a process-related image representing information on the generation process, and displays the image on the display unit 108 . to display. Further, as described above, when a partial region in the document is acquired by the second information acquisition unit 106, the image generation unit 107 generates an image showing the description content of the partial region in the document. It may be generated as a reference image and displayed on the display unit 108 . Note that the reference image is also called a second image. Also, some areas in the document are, for example, paragraphs, as described above. It can also be said that the image generation unit 107 is an output unit that outputs information about the generation process.
  • FIG. 21 is a block diagram showing an example of the functional configuration of the database construction device 200a according to this embodiment.
  • the database construction device 200a analyzes the information in each paper data 1 from the first database collected in advance, and associates the name of one or more raw materials with the generation process corresponding to the one or more raw materials.
  • a second database containing the first generated information 2c shown is constructed. This second database also includes the above-described second generated information 2d.
  • Information in the constructed second database is acquired by the second information acquisition unit 106 of the generation process display device 100a and used by the generation learning unit 104a.
  • the database construction device 200a generates a known material with the material name using one or more raw materials from all paragraphs in the article data 1 to which the material name is assigned in advance. Identify paragraphs that contain content about Then, the database construction device 200a acquires the name of the production process and the name of one or more raw materials in the specified paragraph, and associates the name of the production process with the name of one or more raw materials. generates the first generated information 2c and adds it to the second database. When one or more raw material names are assigned to the article data 1 as material names, the database construction device 200a uses the one or more assigned raw material names to generate the first generated information 2c. may be generated. That is, the database construction device 200a acquires the name of the generation process in the specified paragraph, and associates the name of the generation process with the name of one or more raw materials given to the article data 1. to generate the first generation information 2c.
  • the database construction device 200a uses a pre-learned generation information predictor to identify paragraphs that describe the generation of known materials. Further, the database construction device 200a analyzes the paper data 1 to extract information used by the image generation unit 107. FIG. For example, the database construction device 200a detects and extracts a paragraph describing a generation process from the article data 1. FIG.
  • Such a database construction device 200a includes an article acquisition unit 202, an area division unit 203, an area descriptor generation unit 204, a labeling unit 205, a generated information learning unit 206a, a label acquisition unit 207, and an information extraction unit. 208;
  • the article acquisition unit 202, the area division unit 203, and the area descriptor generation unit 204 perform the same processing as in the first to fifth embodiments.
  • the label assigning unit 205 assigns labels to the paragraphs obtained from the area dividing unit 203 according to the input operation by the user of the database construction device 200a. This label indicates whether the paragraph contains information about the production of the material. Note that in the production of the material, the material is produced from one or more raw materials. Then, the label assigning unit 205 outputs the assigned label to the generated information learning unit 206 a and the label acquiring unit 207 .
  • the generated information learning unit 206a acquires region descriptors corresponding to each of the plurality of paragraphs from the region descriptor generation unit 204, and acquires labels given to each of the plurality of paragraphs from the labeling unit 205. FIG. Then, for each of the plurality of paragraphs, the generated information learning unit 206a learns the relationship between the region descriptor corresponding to the paragraph and the label attached to the paragraph, thereby obtaining a generated information predictor that indicates the relationship. to build. For example, clustering is used as the prediction means, as in the first to fifth embodiments.
  • the generation information predictor outputs a label indicating whether or not the paragraph corresponding to the region descriptor contains information regarding the generation of the material for the region descriptor input.
  • the generation information learning unit 206a acquires the region descriptor from the region descriptor generation unit 204 without acquiring the label from the label assignment unit 205. After obtaining the region descriptor, the generated information learning unit 206a inputs the region descriptor to the generated information predictor, thereby obtaining the aforementioned label corresponding to the region descriptor from the generated information predictor. The generated information learning unit 206 a outputs the label obtained from the generated information predictor to the label obtaining unit 207 .
  • the label obtaining unit 207 obtains a paragraph from the region dividing unit 203, and obtains a label corresponding to the paragraph from the label assigning unit 205 or the generated information learning unit 206a. Before the generated information predictor is constructed, that is, before the generated information predictor is trained, the label obtaining unit 207 obtains a label from the label assigning unit 205 . On the other hand, after the generated information predictor is constructed, that is, after learning of the generated information predictor, the label obtaining unit 207 obtains the label from the generated information learning unit 206a.
  • the label obtaining unit 207 gives the paragraph obtained from the region dividing unit 203 a label corresponding to the paragraph obtained from the label assigning unit 205 or the generated information learning unit 206a.
  • the label acquisition unit 207 then outputs the labeled paragraph to the information extraction unit 208 .
  • the information extraction unit 208 acquires the labeled paragraph from the label acquisition unit 207 .
  • Information extractor 208 then extracts the information from the paragraphs labeled as containing information about the production of the material.
  • the information is information about the production process of the material, and indicates, for example, at least one of the name of the production process, the name of one or more raw materials, the compositional formula, the crystal structure, the characteristic values, and the process conditions.
  • the information extraction unit 208 extracts the name of the generating process from the paragraph by searching for a character string indicating the name of the generating process from the paragraph.
  • the information extraction unit 208 also extracts the name of the raw material from the paragraph by searching for a character string indicating the name of the raw material from the paragraph.
  • the information extraction unit 208 extracts from the paragraph the unit of temperature used in the generation process and the numerical value placed in front of the unit, thereby extracting the process condition indicating the temperature. Further, the information extracting unit 208 extracts information indicating the crystal structure from the paragraph by searching for a character string indicating the crystal structure from the paragraph. Further, the information extracting unit 208 extracts information indicating the characteristic value from the paragraph by searching the unit of the characteristic value and the numerical value placed before the unit from the paragraph.
  • the information extraction unit 208 associates the name of the extracted production process with the name of one or more raw materials. As a result, the information extraction unit 208 generates the first generation information 2c that indicates one or more raw materials for generating the known material and the generation process in association with each other. Furthermore, the information extraction unit 208 associates the attributes of the known materials other than the generation process with the names of one or more raw materials. As a result, the information extracting unit 208 generates the second generated information 2d that indicates one or more raw materials and attributes of known materials in association with each other.
  • 22A and 22B are diagrams showing an example of information indicated by the second database.
  • the second database summarizes the plurality of first generated information 2c and the plurality of second generated information 2d in one tabular format.
  • This second database shows the name of the magnetic material and the name of the resin as the names of the two raw materials. Names of magnetic substances are, for example, "Magnetic nanocomposite " and " Fe3O4 ".
  • the numbers indicating the composition ratios in the composition formula may be written as subscripts, or may be written as ordinary letters that are not subscripts.
  • the magnetic material is a "Magnetic nanocomposite” composed of " Fe3O4 ", "Magnetic nanocomposite " and “ Fe3O4 " are combined like “Magnetic nanocomposite ( Fe3O4 )”.
  • the names of resins are, for example, "PVA", "cellulose” and the like.
  • the second database shows the document name, which is the name of the article data 1, the use of the material, etc., as attributes of the known material.
  • the second database shows the names of production processes such as ultrasonic treatment, freeze-drying, impurity removal, etc., and process conditions such as resin mixing temperature, resin mixing time, magnetic material stirring speed, and melting time as attributes of known materials. show.
  • FIG. 23 is a flow chart showing an example of the overall flow of processing of the generated process search system 10a according to the present embodiment.
  • Step S1100a The database construction device 200a constructs a second database including first generation information 2c and second generation information 2d indicating names of one or more raw materials and production processes in association with each other.
  • Step S1200a The production process display device 100a builds a production predictor that predicts the production process from the name of one or more ingredients by performing learning using the first production information 2c.
  • a fully-connected NeuralNetwork is used for the model of the generating predictor.
  • Step S1300a The production process display device 100a uses a production predictor to derive production probabilities for each type of production process for producing a target material from one or more raw materials. That is, the production process display device 100a predicts, for each of a plurality of types of production processes, a production probability indicating the degree to which that type of production process is required to produce a target material using one or more raw materials.
  • FIG. 24 is a flowchart showing an example of the overall flow of processing by the database construction device 200a according to this embodiment. 24 is a flowchart showing the detailed flow of step S1100a in FIG.
  • each of the plurality of paper data 1 stored in the first storage unit 301 contains a magnetic resin material, which is a composite material of a magnetic substance (specifically, magnetic powder) and a resin. name is associated.
  • the article data 1 and the material name are linked by manually confirming the material name mainly handled in the article data 1 and assigning it to the article data 1 .
  • the method of assigning the material name is not limited to this, and may be automatic.
  • the magnetic resin material may also be called a magnetic resin mixed material, a magnetic powder resin material, or a magnetic powder resin mixed material.
  • Step S1101a The paper acquisition unit 202 acquires one paper data 1 to be processed from the first database containing a plurality of paper data 1 already published, and outputs the paper data 1 to the area division unit 203 . That is, the paper data 1 to which the material name is linked in advance is selected.
  • Step S1102 The region division unit 203 divides the article data 1 acquired from the article acquisition unit 202 into a plurality of paragraphs and acquires them.
  • the region division unit 203 outputs each obtained paragraph to the label acquisition unit 207 , the region descriptor generation unit 204 , and the label assignment unit 205 .
  • Step S1103a Upon obtaining a paragraph from the region dividing unit 203, the region descriptor generating unit 204 generates a region descriptor corresponding to the paragraph.
  • the production information learning unit 206a inputs the region descriptor to the production information predictor so that the paragraph corresponding to the region descriptor contains information on the production of the material of the material name. Gets a label indicating whether the is included. Then, generated information learning section 206 a outputs the label to label acquisition section 207 .
  • the labeling unit 205 Before construction of the generated information predictor, the labeling unit 205 outputs the above label to the label acquisition unit 207 according to the input operation by the user of the database construction device 200a.
  • the label acquisition unit 207 acquires the label output from the generation information learning unit 206a or the labeling unit 205, that is, the label indicating whether or not the paragraph describes the generation of materials, and adds the acquired label to the paragraph. do.
  • the label acquisition unit 207 then outputs the labeled paragraph to the information extraction unit 208 .
  • Step S1104a The information extraction unit 208 extracts the name of one or more raw materials and the production process from the paragraphs describing the production of ingredients, i.e., the paragraphs containing information on the production of ingredients, among the paragraphs acquired from the label acquisition unit 207. Extract information indicating the name, etc. A plurality of names of raw materials, a plurality of names of production processes, and the like are registered in advance in the information extraction unit 208 .
  • the information extraction unit 208 determines whether or not the name of the registered production process exists in the paragraph describing the production of the material. When the information extraction unit 208 determines that there is a name of the registered generation process, it extracts the name of the generation process.
  • the information extracting unit 208 extracts the names of all those generated processes.
  • the information extraction unit 208 also extracts the name of one or more raw materials as well as the name of the production process.
  • the name of the one or more raw materials used to produce the known material includes the name of the magnetic substance and the name of the resin.
  • the information extracted by the information extraction unit 208 indicates not only the name of the generation process and the names of one or more raw materials, but also the attributes of known materials included in the article data 1 acquired in step S1101a.
  • This attribute is the document information of the paper data 1, the location of the paragraph that describes the name of the production process, or the crystal structure, process conditions, characteristic values, or use of the known material described in the paragraph and so on.
  • Step S1105a The information extraction unit 208 further associates the name of the production process extracted in step S1104a with the name of one or more raw materials, and adds them to the second database. That is, the first generated information 2c is generated and added to the second database.
  • the information extracting unit 208 also associates attributes other than the generation process extracted in step S1104a with the names of one or more raw materials, and adds them to the second database. That is, the second generated information 2d is generated and added to the second database.
  • FIG. 25 is a diagram showing an example of a process-related image according to this embodiment.
  • the image generation unit 107 generates a process-related image Pa1 and outputs it to the display unit 108.
  • FIG. This process-related image Pa1 includes a raw material area da1, a known material area da2, and a production possibility area da3.
  • the raw material area da1 the name of one or more raw materials for generating the target material indicated by the first information acquired by the first information acquisition unit 102 is displayed.
  • the raw material area da1 includes a first raw material area da11 and a second raw material area da12.
  • the production possibility area da3 displays the production probabilities derived by the probability derivation unit 105, that is, the production probabilities of each of multiple types of production processes for producing the target material using one or more raw materials. That is, the process-related image Pa1 displays the names of one or more raw materials for producing the target material, which are input by the user, and the production probability for each type of production process for producing the target material.
  • the name of the magnetic substance " Fe3O4 " and the name of the resin "PVA” are respectively displayed. These names were entered by the user.
  • the magnetic material is used as magnetic powder. Composite materials of magnetic powder and resin are produced as target materials from raw materials with these names.
  • the name of the raw material may be a composition formula such as " Fe3O4 ", a registered trademark such as " LAPONITE ", or a chemical formula such as "magnetite”. It may be a product name.
  • the name of the raw material may be an abbreviation such as “PVA” (or “PVAL”, “PVOH”, etc.), an abbreviation such as “Poval”, or “polyvinyl alcohol ( It may be a chemical product name such as "polyvinyl alcohol)”.
  • the production probability for each type of production process for producing the composite material, which is the target material, from the magnetic material "Fe3O4" and the resin "PVA" is displayed.
  • This production probability is an example of the production parameter described above and indicates the degree to which the production process is required to produce the target material.
  • a production probability or production parameter may be referred to as production probability, which is the possibility that the production will take place, or feasibility, which is the possibility that the production will take place.
  • the generation possibility area da3 includes an object da31 indicating the name of each of a plurality of types of generation processes, and a probability frame da32 for displaying the generation probability of the generation process with the name indicated by each of these objects da31.
  • the types of production processes in the present embodiment include seven types of resin material stirring, magnetic material/resin mixing, ultrasonic treatment, freezing, melting, freeze-drying, and impurity removal, but are not limited to these. do not have.
  • the type of generated process displayed is based on the first generated information 2c. For example, in the probability frame da32 corresponding to the ultrasonically processed object da31 in the production possibility area da3, the probability that the ultrasonic treatment is required to produce the composite material "60%" is displayed as the production probability.
  • these objects da31 are displayed so as to be selectable by the user. That is, the user can select one of the objects da31 by performing an input operation on the input unit 110. FIG. In other words, the user can select the desired object da31, ie the generation process.
  • the known material area da2 As an example, a subarea having tabs related to the production process selected by the user is displayed.
  • the known material area da2 has two sub-areas each with a tab.
  • information is displayed as secondary information about one or more relevant raw materials used to create the known material for the creation process selected by the user.
  • information about one or more relevant raw materials used to produce the known material without the production process selected by the user is displayed as secondary information.
  • the second information indicates, for example, the names of two related raw materials and the literature information of the paper data 1 describing the production process of the known material using those related raw materials.
  • the display of the two sub-areas is switched by selection of those tabs by the user.
  • not only one piece of second information but also a plurality of pieces of second information may be displayed in the sub-region. That is, in the sub-region, for each of the one or more known materials, information regarding the production of each of the one or more known materials is displayed as the second information.
  • the image generator 107 generates, for each type of generation process, a process-related image indicating second information about each of one or more known materials generated by performing that type of generation process. Generate Pa1. This allows the user, for example, to provide process-related second information for each of one or more known materials produced using the same production process required to produce a destination material using two raw materials. It can be easily found out from the image Pa1.
  • the image generation unit 107 in the present embodiment generates a process-related image Pa1 indicating second information about each of one or more known materials generated without performing the type of generation process for each type of generation process. Generate. This allows the user to easily obtain, from the process-related image Pa1, second information for each of one or more known materials that are produced without a production process that is not required to produce a target material using, for example, two raw materials. can be found in
  • the sub-region having a tab describing "with ultrasonic processing” is displayed first.
  • the tab "with ultrasonic processing” is a tab relating to generation using the generation process "ultrasonic processing” selected by the user from among the seven types of generation processes. That is, the image generation unit 107 first outputs to the display unit 108 a sub-region having tabs relating to generation using the generation process selected by the user.
  • the generation probability of the generation process “sonication” selected by the user is 60%.
  • the image generator 107 first displays a sub-region having a tab related to generation using the generation process “ultrasonic processing” selected by the user. good.
  • the image generator 107 may first display a sub-region with a tab for production without the production process “sonication” selected by the user. good. That is, a subregion with a tab stating "no sonication" may be displayed first.
  • the image generation unit 107 displays the sub-region having tabs related to generation using the generation process selected by the user and the sub-region having tabs related to generation not using the generation process selected by the user side by side. may be displayed in This allows the user to easily visually grasp known information regarding generation using the selected generation process and generation not using the selected generation process.
  • a combination of one or more related raw materials is hereinafter also referred to as a related raw material set, and a combination of one or more raw materials indicated by the first information is hereinafter also referred to as a raw material set.
  • a related raw material set is a set of materials similar to the raw material set, and is a combination of one or more materials whose vector distances are close to the raw material set. That is, a related ingredient set is a set of ingredients that have descriptors close to those of the ingredient set.
  • the line includes the name of the magnetic substance "Fe 3 O 4 " and the name of the resin "PVA (+DMSO)".
  • the predetermined order is the order of shortest distance between vectors.
  • the vector of the related raw material group consisting of the magnetic substance " Fe3O4 " and the resin "PVA ( +DMSO)" is the vector of the raw material group consisting of the magnetic substance " Fe3O4 " and the resin "PVA”.
  • the composite vector obtained from the respective vectors of the magnetic material " Fe3O4 " and the resin "PVA ( +DMSO)" is the name of the magnetic material " Fe3O4 " and the name of the resin "PVA” indicated in the first information. , which is closest to the composite vector obtained from each vector of . Therefore, a line including the name of the magnetic substance " Fe3O4 " and the name of the resin "PVA ( +DMSO)" is displayed at the top.
  • the order of rows including the names of one or more related raw materials may be determined by, for example, the second information acquisition unit 106 and notified to the image generation unit 107 .
  • the second information of the related raw material set having the distance between vectors equal to or less than the threshold is displayed.
  • the second information of each of the predetermined number of related raw material sets may be displayed in order of shortest distance between the above vectors.
  • the image generation unit 107 in this embodiment arranges the second information of each of one or more related raw material sets in an order according to the vectors or descriptors of the related raw material sets, and arranges the second information in that order.
  • a process-related image Pa1 is generated that indicates the one or more pieces of second information that have been obtained.
  • the second information indicating a related raw material set that is identical or similar to the raw material set for producing the target material is placed first. Thereby, the user can easily find out the desired second information from the second information.
  • the two related raw materials displayed in the known material area da2 are the same as the two raw materials for producing the target material. That is, the two related raw materials, the magnetic substance “Fe 3 O 4 ” and the resin “PVA (+DMSO)”, and the two raw materials shown in the first information, the magnetic substance “Fe 3 O 4 ” and the resin “PVA” is the same as Further, in the known material area da2, it is indicated that the production process "sonication" selected by the user is used to produce the known material using the two related raw materials. In addition, in the name of the related raw material indicated by the second information, the part that differs from the name of the raw material indicated by the first information may be displayed in a more emphasized manner than the same part. For example, the different parts may be emphasized, such as by bold. In the example of FIG. 25, "DMSO", which is described as the solvent for the resin "PVA”, is highlighted in bold.
  • the attribute of the known material in the second information displayed in the known material area da2 includes, for example, the document information of the paper data 1 that is the source of the known material.
  • Literature information that serves as a source of known materials is obtained from the second generated information 2d of the second database.
  • the literature information may be acquired from other databases via network 401, such as the Internet.
  • FIG. 26 is a diagram showing another example of the process-related image Pa1 according to this embodiment.
  • the generation probabilities of seven types of generation processes are displayed.
  • the process-related image Pa1 shown in FIG. 26 only the generation probabilities of the type of generation process selected by the user among the seven types of generation processes are displayed. That is, only the probability frame da32 corresponding to the object da31 selected by the user is displayed without displaying the probability frame da32 for each of the seven objects da31 each describing the name of the generation process.
  • the probability frame da32 displays the generation probability of the generation process having the name described in the object da31 selected by the user. For example, the generation probability "60%" of the generation process "ultrasonic processing" is displayed in the probability frame da32.
  • the user can focus only on the generation probability of a specific generation process that they want to consider whether it is necessary or not to generate the target material. Therefore, it is possible to display only the generation probability of the generation process that the user wants to study, and to optimally support material development.
  • FIG. 27 is a flow chart showing an example of the overall flow of processing of the generating process display device 100a according to the sixth embodiment. 27 is a flowchart showing the detailed flow of step S1300a in FIG.
  • the first input unit 101 receives names of one or more raw materials to be displayed by user's input operation, and outputs the names to the first information acquisition unit 102 .
  • Step S1302a The first information acquisition unit 102 acquires the names of one or more ingredients from the first input unit 101 and outputs them to the descriptor generation unit 103a.
  • Step S1303a The descriptor generation unit 103 a generates descriptors corresponding to the names of the one or more raw materials acquired from the first information acquisition unit 102 and outputs the descriptors to the probability derivation unit 105 and the second information acquisition unit 106 .
  • Step S1304a The generation learning unit 104a inputs the one or more descriptors acquired from the descriptor generation unit 103a through the probability derivation unit 105 to the trained generation predictor, thereby performing a generation process corresponding to the one or more descriptors. Get the generation probability for each type of . Then, generation learning section 104 a outputs the generation probability to probability derivation section 105 . Note that when the second database is updated, the generation learning unit 104a may re-learn the generation predictor based on the difference between the post-update and pre-update.
  • the probability derivation unit 105 obtains the generation probability for each type of generation process from the generation learning unit 104a, and calculates the generation probability for each type of generation process for generating the target material using one or more raw materials. derive The probability derivation unit 105 outputs the generation probability for each type of generation process to the image generation unit 107 .
  • Step S1305a The image generation unit 107 generates a process-related image Pa1 indicating the name of one or more raw materials, which is input by the user and is acquired from the first information acquisition unit 102 .
  • This process-related image Pa1 shows seven types of generation processes and the generation probabilities for each type of the generation processes obtained from the probability deriving unit 105 .
  • this process-related image Pa1 includes the second information of each of the one or more related raw material groups related to the raw material group, acquired from the second information acquisition unit 106 .
  • the second information indicates, for example, the names of one or more related raw materials included in the related raw material group and the attributes of known materials generated from the one or more related raw materials. Attributes include document information, the location of paragraphs, and the like.
  • Step S1306a The image generation unit 107 displays the process-related image Pa1 generated in step S1305a on the display unit .
  • Step S1307a For example, the first information acquisition unit 102 selects a type of generation process according to the user's input operation on the first input unit 101 from among the seven types of generation processes shown in the process-related image Pa1. The first information acquisition unit 102 then notifies the image generation unit 107 of the selected generation process.
  • Step S1309a The image generation unit 107 displays one or more pieces of second information related to the notified generation process, ie, the generation process selected by the user, in the known material area da2 of the process-related image Pa1.
  • the second information indicates the name of one or more relevant raw materials for producing the known material using the production process and attributes of the known material.
  • the second information indicates the name of one or more relevant raw materials for producing the known material without using the production process and attributes of the known material.
  • the first information indicates not only one or more raw materials for producing the target material, but also the function of the target material.
  • descriptor generator 103a generates a descriptor of the function indicated by the first information.
  • the probability derivation unit 105 derives a production probability for each type of production process for producing a target material having the function from the one or more raw materials. That is, the probability derivation unit 105 acquires the descriptors of each of the one or more ingredients and the descriptors of their functions from the descriptor generation unit 103a, and outputs these descriptors to the generation learning unit 104a.
  • the generation learning unit 104 a obtains generation probabilities for each type of generation process corresponding to the descriptors by inputting the descriptors into the generation predictor, and outputs the generation probabilities to the probability derivation unit 105 . do.
  • the probability derivation unit 105 derives these generation probabilities by acquiring those generation probabilities from the generation learning unit 104a.
  • the production process for each type of production process for producing at least one of the target material with arbitrary function and the target material without any function Generation probabilities are derived.
  • the first generation information 2c indicates not only the name of the generation process but also the name of the function of the known material generated from one or more raw materials in association with the name of the one or more raw materials.
  • the production process is a process for producing a known material having the function from the one or more raw materials.
  • the generation predictor of the generation learning unit 104a is constructed by learning using such first generation information 2c.
  • the information extraction unit 208 of the database construction device 200a in this modified example also extracts the names of the functions of known materials from the paragraphs of the paper data 1 in order to generate the above-described first generated information 2c.
  • the information extractor 208 extracts not only the name of one or more raw materials used in the production of the known material, the name of the production process, and the attributes of the known material, but also the names of the functions of the known material.
  • each article data 1 in the first database may be given not only the material name described above but also the function name of the known material.
  • the information extraction unit 208 generates the first generated information 2c indicating the name of the function assigned to the article data 1 in association with the name of one or more raw materials.
  • the function of the known material may be indicated in the second database as an attribute of the known material.
  • the functional descriptor generated by the descriptor generation unit 103a may be described as the above-described One Hot Vector. Functionality is indicated by information (also called categorical information) that is not suitable to be represented as physically continuous variables, such as temperature, time, and so on.
  • the functions of the magnetic resin material include adsorption and conductivity.
  • a vector [0, 0] with the number of dimensions "2" is prepared for each of adsorption and conductivity.
  • Descriptors for the feature 'adsorption' are represented as vectors [1,0] and descriptors for features 'conductivity' are represented as vectors [0,1].
  • the element corresponding to the object (ie function) indicated by the descriptor is flagged, that is, the value of the corresponding element is changed from 0 to 1.
  • the number of dimensions is four, and when there are six function candidates, the number of dimensions is six. In this way, the number of dimensions (in other words, the length) of the vector is set according to the number of function candidates.
  • FIG. 28 is a diagram showing an example of the process-related image Pa1 according to this modified example.
  • check boxes d6 for selecting each of a plurality of functions are displayed.
  • the user first performs an input operation on the first input unit 101 to input the name of the magnetic substance " Fe3O4 " and the name of the resin "PVA" into the raw material area da1 of the process-related image Pa1.
  • the user selects the object da31 describing “ultrasonic processing” as the generation process by performing an input operation on the first input unit 101 .
  • the generation probability "60%" of the generation process "ultrasonic processing” is displayed in the probability frame da32.
  • the user selects, for example, one function from the plurality of functions by performing an input operation on the first input unit 101 .
  • the user inputs a check mark in any of the check boxes d6 associated with each of the plurality of functions.
  • the function associated with the check box d6 with the check mark is selected.
  • Each of the multiple functions is a function that the target material has, such as adsorption, conductivity, and heat resistance.
  • Adsorption is the ability of the target material (ie, magnetic resin material) to adsorb chemicals, drugs, and the like.
  • Conductivity is the ability of a target material to conduct electricity.
  • Heat resistance is a function of the target material being able to withstand heat at high temperatures.
  • the user selects the function "adsorption" among those functions.
  • the first information indicates not only the names of the two ingredients previously entered by the user, but also the name of the selected function.
  • the probability derivation unit 105 updates the generation probabilities of the seven types of generation processes based on the first information. For example, the probability deriving unit 105 updates the generation probability of the generation process “ultrasonic processing” from “60%” to “90%”. The generation probability updated in this way is displayed in the probability frame da32 as shown in FIG.
  • the target materials generated using the generation process can be narrowed down to target materials having functions selected by the user.
  • the production probabilities of a production process can be narrowed down to the probability that the production process is required to produce a target material having the functionality selected by that user.
  • the generation probability of the generation process "ultrasonic treatment” is updated from "60%" to "90%". . That is, in order to produce a composite material with an adsorption function, there is a high probability that ultrasonic treatment will be required compared to producing a composite material that is not limited in function.
  • the composite material is a target material, more specifically a magnetic resin material. In this way, it is possible to easily and accurately understand the production probability that depends on the function of the target material. This makes it possible to more optimally support material development.
  • the production probability of a production process is expressed as the probability that the production process is required to produce the selected multifunctional target material.
  • FIG. 29 is a diagram showing another example of the process-related image Pa1 according to this modified example.
  • the function is selected by inputting to the check box d6, but as shown in FIG. 29, the function may be selected by selecting options from the pull-down d61.
  • the pull-down d61 is also called a pull-down menu, pull-down list, dropdown, or the like.
  • the option is an item in which the name of the function is written.
  • the generation probability of the generation process selected by the user can be narrowed down to the probability that the generation process is required to generate the target material having the function selected by the user. can be done. Also, even when there are many functions to choose from, the user can easily select a desired function.
  • the function of the target material is selected using the pull-down d61, but the user may directly input the name of the function of the target material.
  • the first information may indicate not only one or more raw materials for producing the target material, the function of the target material, but also the application of the target material.
  • the descriptor generator 103a generates a descriptor for the purpose indicated by the first information.
  • the One Hot Vector described above may be used to generate the descriptor.
  • the probability derivation unit 105 derives a production probability for each type of production process for producing a target material having the function and application from the one or more raw materials.
  • the probability derivation unit 105 acquires descriptors for each of the one or more raw materials and descriptors for their functions and uses from the descriptor generation unit 103a, and outputs these descriptors to the generation learning unit 104a. do.
  • the generation learning unit 104 a obtains generation probabilities for each type of generation process corresponding to the descriptors by inputting the descriptors into the generation predictor, and outputs the generation probabilities to the probability derivation unit 105 . do.
  • the probability derivation unit 105 derives these generation probabilities by acquiring those generation probabilities from the generation learning unit 104a.
  • the type of production process for producing at least one of the target material with any use and the target material without any use Generation probabilities are derived.
  • the first generated information 2c includes the names of one or more raw materials, the names of the functions and uses of known materials generated from the one or more raw materials, and the functions and uses of the known materials.
  • the name of the production process for producing the material from its one or more raw materials is shown in association.
  • the generation predictor of the generation learning unit 104a is constructed by learning using such first generation information 2c.
  • the information extracting unit 208 of the database construction device 200a in this modified example also extracts the names of the functions and the names of uses of the known materials from the paragraphs of the paper data 1 in order to generate the above-mentioned first generated information 2c.
  • the information extraction unit 208 extracts not only the name of one or more raw materials, the name of the production process, and the attributes of the known material used to produce the known material, but also the name of the function and the name of the application of the known material. .
  • each article data 1 in the first database may be assigned not only the material name described above but also the function and application name of the known material.
  • the information extracting unit 208 generates the first generated information 2c that indicates the name of the function and application assigned to the article data 1 in association with the name of one or more raw materials.
  • the functions and uses of the known materials may be indicated in the second database as attributes of the known materials.
  • FIG. 30 is a diagram showing another example of the process-related image Pa1 according to this modified example.
  • check boxes d7 for selecting each of a plurality of uses are also displayed.
  • a user selects, for example, one use from among a plurality of uses by performing an input operation on the first input unit 101 . That is, the user inputs a check mark in any of the check boxes d7 associated with each of the plurality of usages.
  • the application associated with the check box d7 with the check mark is selected.
  • Each of the multiple uses is an application of the target material, such as an adsorbent, a biomaterial, a device material, and the like.
  • Adsorbents are used for adsorption of chemical substances, drugs and the like.
  • Biomaterials are used for purposes such as transplantation into living organisms such as humans.
  • Device materials are applications used for incorporation into machines and the like.
  • the user first performs an input operation on the first input unit 101 to input the name of the magnetic substance " Fe3O4 " and the name of the resin "PVA” into the raw material area da1 of the process-related image Pa1. . Furthermore, the user selects the object da31 describing “ultrasonic processing” as the generation process by performing an input operation on the first input unit 101 . As a result, as shown in FIG. 26, the generation probability "60%" of the generation process "ultrasonic processing” is displayed in the probability frame da32.
  • the user selects, for example, the function "adsorption” from among the plurality of functions, and further selects, for example, the application "adsorption” from among the plurality of uses. (adsorbent)”.
  • the first information indicates not only the names of the two raw materials previously entered by the user, but also the selected function name and use name.
  • the probability derivation unit 105 updates the generation probabilities of the seven types of generation processes based on the first information. For example, the probability deriving unit 105 updates the generation probability of the generation process “ultrasonic processing” from “60%” to “90%”. The generation probability updated in this way is displayed in the probability frame da32 as shown in FIG.
  • the target materials generated using the generation process can be narrowed down to target materials having functions and applications selected by the user.
  • the production probabilities of a production process can be narrowed down to the probabilities that the production process is required to produce a target material having the function and application selected by that user.
  • the generation probability of the generation process "ultrasonic treatment” is "60%". to "90%”.
  • the probability that ultrasonic treatment is required compared to producing a composite material that is not limited in function and application becomes higher. In this way, it is possible to easily and accurately understand the production probability that depends on the function and application of the target material. This makes it possible to more optimally support material development.
  • the function and usage are selected, but only one of the function and usage may be selected. If only an application is selected, the target materials produced using the production process can be narrowed down to those that have the application selected by the user. Thus, the production probability of a production process can be narrowed down to the probability that the production process is required to produce a target material having an application selected by that user.
  • Embodiment 6 a plurality of raw materials are input to the raw material area da1.
  • the second information acquisition unit 106 acquires the second information indicating the raw material having the same or similar name as the name of one raw material indicated by the first information as the related raw material to the second database of the second storage unit 302a. Get from In other words, the second information acquiring unit 106 acquires the second information indicating, as the related raw material, the raw material whose vector-to-vector distance from one raw material indicated by the first information is equal to or less than the threshold.
  • one raw material indicated by the first information is a resin
  • second information is obtained that indicates, as a related raw material, a resin whose vector distance from the resin is equal to or less than a threshold.
  • the second information may indicate not only the resin, which is the related raw material, but also the magnetic material.
  • the magnetic material is any magnetic material that is not restricted by the first information.
  • 31 and 32 are diagrams showing an example of the process-related image Pa1 according to this modified example.
  • the user does not enter the name of the raw material, which is a magnetic substance, in the first raw material area da11, and enters the name of the raw material, which is resin, "cellulose" in the second raw material area da12.
  • the production process display device 100a obtains the first information indicating the raw material name "cellulose”, and derives the production probabilities of the seven types of production processes based on the first information.
  • the generation process display device 100a displays the generation probability " 10%” is displayed in the probability frame da32.
  • the user does not enter the name of the raw material, which is a magnetic material, in the first raw material area da11, and enters the name of the raw material, which is resin, "PVA" in the second raw material area da12.
  • the production process display device 100a obtains the first information indicating the name of the raw material "PVA”, and derives the production probability of each of the seven types of production processes based on the first information.
  • the generation process display device 100a displays the generation probability " 90%” is displayed in the probability frame da32.
  • second information indicating related raw materials having the same or similar name as the raw material name "cellulose", which is a resin, input by the user is displayed.
  • the second information indicating the name "cellulose" of the related raw material, which is resin is displayed.
  • the second information indicates the name of the magnetic material "Fe 3 O 4 (+BzMe3NOH)" and the attributes of known materials produced by the magnetic material and resin.
  • the attributes include the document information of the article data 1 describing the production of the known material, and the conditions (that is, process conditions) used for the production process "resin material stirring" selected by the user.
  • the second information indicating the names of the related raw materials which are resins, "PVA (+DMSO),""PVA,” and “PVA (+distilled water),” respectively, is displayed.
  • the second information indicating the name of the related raw material "PVA (+DMSO)” includes not only the name but also the name of the magnetic substance " Fe3O4 ( +BzMe3NOH)" and the name of the magnetic substance and the resin. properties of known materials.
  • the second information indicating the name of the related raw material "PVA” indicates not only the name but also the name of the magnetic substance "nano-magnetite” and the attributes of the known material produced by the magnetic substance and the resin. .
  • the second information indicating the name of the related raw material "PVA (+distilled water)” includes not only the name but also the name of the magnetic material "LAPONITE-RD” and the attributes of the known material produced by the magnetic material and the resin. show. These attributes include the literature information of article data 1 describing the production of the known material, and the process conditions used for the production process "stir resin material” selected by the user, as described above. For example, when a known material is produced from the magnetic material " Fe3O4 ( +BzMe3NOH)" and the resin "PVA (+DMSO)", the process condition of the production process "resin material stirring” is 6 hours of stirring at 80°C. is shown in the second information as
  • the related raw materials of the magnetic material which are the raw materials that have not been input in the raw material area da1
  • the process conditions according to the second information displayed in the known material area da2.
  • the production of a magnetic resin material which is an example of an organic/inorganic mixed material
  • the number of days required for the generation is displayed as the process condition.
  • the user can have some idea of the production processes that are feasible in the user's situation based on the process conditions before embarking on experimentation with the production of the target material, greatly increasing the efficiency of the production process search. can be improved.
  • equipment used in the generation process may be displayed as the process conditions. For example, different production processes may be required depending on the raw materials used, and different equipment is required depending on the production processes. By displaying such equipment, the user can easily determine that the target material can be generated by the generation process if the user owns the equipment.
  • the function and application of the target material may be input. Even if the magnetic resin material is the same, if the function and application are different, the production process required to produce the magnetic resin material may be different, and the necessary equipment is also different depending on the production process. In this modified example, even in such a case, by displaying the necessary equipment, if the user owns the equipment, the user can generate the magnetic resin material by the generation process. can be easily determined.
  • FIG. 33 is a diagram showing another example of the process-related image Pa1 according to this modified example.
  • the process conditions are displayed as numerical values in the example of FIG. 32, they may be displayed as a map ma as shown in FIG.
  • the image generation unit 107 may generate a map ma indicating the attribute relationship of the known materials indicated by the second information of each of one or more known materials, and output the map ma to the display unit 108 .
  • the known material is a composite material (that is, a magnetic resin material) produced from a magnetic material and a resin, which are related raw materials.
  • "resin material stirring” is selected as the generation process.
  • a map ma showing the relationship between the stirring time and stirring speed used as the process conditions of the generation process and the number of article data 1 describing the stirring time and stirring speed is displayed.
  • the horizontal axis of the map ma indicates the stirring speed
  • the vertical axis indicates the stirring time.
  • the color or color depth corresponds to the number of paper data 1 describing the stirring speed and stirring time. It is shown. For example, the greater the number of article data 1, the darker black is displayed, and the lesser the number, the lighter black is displayed.
  • the user can easily visually compare the attributes of the known materials indicated by the second information of each of the one or more known materials. This makes it possible to more optimally support material development.
  • the name of the resin is entered in the raw material area da1 and the name of the magnetic substance is not entered. Conversely, the name of the magnetic substance is entered and the name of the resin is not entered. may Even in such a case, the same effect as described above can be obtained.
  • FIG. 34 is a diagram showing an example of the process-related image Pa1 according to this modified example.
  • the image generation unit 107 displays on the display unit 108 a process-related image Pa1 showing only generation processes corresponding to generation probabilities of, for example, a threshold of "50%" or more among the seven types of generation processes. For example, the image generation unit 107 does not display the object da31 indicating the generation process "resin material stirring" among the seven types of generation processes, and displays only the objects da31 of the other six types of generation processes.
  • the user does not enter the name of the raw material that is a magnetic material in the first raw material area da11, and enters the name of the raw material that is resin in the second raw material area da12. input.
  • the generation probability of the generation process “resin material stirring” is 10%, which is equal to or less than the threshold “50%”. Therefore, the object da31 indicating this generation process "stir resin material" is not displayed.
  • FIG. 35 is a diagram showing another example of the process-related image Pa1 according to this modified example.
  • the threshold "50%" is predetermined.
  • the threshold can be arbitrarily set by the user. That is, the image generation unit 107 displays the process-related image Pa1 including the threshold frame d8 into which the threshold is input on the display unit 108 . The user inputs a threshold in the threshold frame d8 by performing an input operation on the input unit 110. FIG. The user may input a condition such as greater than or equal to the threshold value or less than the threshold value in the threshold frame d8. As a result, only the generation process that satisfies the threshold condition input by the user is included in the process-related image Pa1 and displayed.
  • the generation probability is not displayed for each of the displayed generation processes, but the generation probability may also be displayed as in the example shown in FIG. .
  • Embodiment 6 and Modifications 1 to 3 have been described above, the present disclosure is not limited to Embodiment 6 and Modifications 1 to 3 thereof. Various modifications conceived by those skilled in the art may be included in the present disclosure as long as they do not deviate from the gist of the present disclosure. Also, the sixth embodiment and its modifications 1 to 3 may be combined with any of the first to fifth embodiments.
  • the names of two raw materials are entered in the raw material area da1, but the names of three or more raw materials may be entered.
  • the raw material is not limited to the magnetic material and the resin, and may be any material.
  • the generation probability of each of a plurality of generation processes is derived in the sixth embodiment and its modifications 1 to 3, the generation probability of one generation process may be derived.
  • the generating process display devices 100 and 100a according to the first to sixth embodiments and their modifications are the following generating process output devices.
  • the production process output device includes a first information acquisition unit 102 that acquires first information about the target material, and for each of one or more types of production processes, the type of target material specified by the first information.
  • the output unit is, for example, the image generation unit 107 that generates and outputs an image including generation parameters.
  • the first information may directly indicate the target material, such as the composition of the target material, or indirectly indicate the target material, based on one or more raw materials used to produce the target material.
  • the output unit outputs a generation parameter indicating at least one of a generation probability and a rank indicating the above-mentioned degree, and the generation probability indicates that a type of generation process corresponding to the generation probability is required to generate the target material.
  • the order is the order in which each of the one or more generation processes is ranked in the order of degree mentioned above.
  • the derivation unit described above is, for example, the probability derivation unit 105 that derives the generation probability.
  • the order is, for example, the order shown in FIG. 8A.
  • the database is, for example, the second database in the first to sixth embodiments.
  • the production parameters for each type of production process for the target material are output. That is, for each type of generation process, the user can easily grasp the necessity of the generation process for the generation of the target material, in other words, the possibility of generating the target material by the generation process. Therefore, when generating a target material, the user can select a generation process after grasping the possibility of generation for each type of generation process. As a result, even if the user is a researcher with little knowledge and experience, he or she can easily select the generation process. Also, if the user is a researcher with a wealth of knowledge and experience, the user's knowledge and experience can be used to select a more accurate generation process.
  • the user finds the type of generation process that is most required for the generation of the target material from those generation processes. be able to. That type of production process can also be said to be the type of production process most likely to produce the desired material. The user can then determine that the production process can be used to produce the target material. As a result, the user's search for the production process of the target material can be more appropriately supported.
  • the above-mentioned derivation unit in response to the input of the descriptor related to the material, the predictor learned using the database so that each generation parameter of one or more types of generation processes for the material is output, Inputting a descriptor for a target material derives production parameters for each of one or more production processes for that target material.
  • the predictor is, for example, the baking predictor or production predictor described above, and in specific examples, a neural network or the like.
  • the predictor outputs the respective generation parameters of one or more types of generation processes for the material having the composition for the input of the descriptor indicating the composition.
  • the derivation unit inputs a descriptor indicating the composition of the target material to the predictor, thereby making each of the one or more types of generation processes for the target material derive the generation parameters of
  • the production probabilities for each of the plurality of types of baking methods are derived as the production parameters for each of the plurality of types of production processes.
  • the predictor uses a database such that for input of descriptors indicating one or more raw materials for producing a material, production parameters for each of the one or more production processes for the material are output.
  • the derivation unit inputs into the predictor descriptors indicative of one or more raw materials for producing the target material, thereby generating the one or more production processes for the target material.
  • the generation parameters for each of For example, as in Embodiment 6, generation probabilities for each of a plurality of types of generation processes are derived as generation parameters.
  • the production parameters of each of the one or more production processes for the unknown target material can be obtained by using the learned predictor. can be derived.
  • the output unit also outputs the generation parameters by generating a first image showing at least one generation parameter and outputting the first image to the display unit 108 . That is, the output unit in this case corresponds to the image generation unit 107 . Also, the first image is, for example, the process-related images P1 and Pa1 described above.
  • the first image showing the generation parameters is displayed on the display unit 108, so that the user can easily comprehend the generation parameters by viewing the first image.
  • the user can generate the target material using the type of generation process corresponding to the largest generation parameter. can be determined.
  • the user confirms the generation parameters corresponding to the generation processes possible with the generation device that is owned. be able to.
  • the user may try to generate the target material by the generation process even if the generation process corresponds to the smallest generation parameter, if the generation parameter indicates a generation probability of 30% or more, for example. judgment can be made.
  • the production process output device like the production process display device 100 of Embodiments 1 to 5, for each of one or more known materials related to the target material, outputs the second information indicating the known material to the material and the output unit generates a first image further indicative of second information for each of the one or more known materials, the second information obtainer 106 obtaining the first image
  • the output unit may generate an image showing the description content of the partial region in the document as a second image and output it to the display unit 108 .
  • the second information indicating the known material and the partial area of the document describing the production process of the known material are displayed.
  • the user can select a production process after grasping the possibility of production for each type of production process from a broader point of view when producing a target material.
  • the user can easily grasp what kind of production process is being performed for a known material similar to the target material. Therefore, it is possible to more appropriately support the search for the production process of the target material.
  • each generation parameter of N types of generation processes (N is an integer of 3 or more) is derived by the derivation unit, and among the N types of generation processes, the degree of the above is greater than the remaining other generation processes
  • the output unit outputs M types of generation processes among the N types of generation processes.
  • a first image is generated that shows the production parameters of each of the processes and not the production parameters of the remaining other production processes.
  • the predetermined condition is that the generation probability is equal to or greater than the threshold value of 80%. Therefore, the user can easily find the most likely production process for the target material. Further, for example, if the sum of the generation probabilities of the top two of the generation probabilities of three or more types of generation processes is equal to or greater than the threshold value of 90%, only the top two generation probabilities are displayed, and the remaining other generation probabilities are displayed.
  • the top two generation probabilities are two generation probabilities that are higher than any other generation probabilities.
  • the predetermined condition is that the sum of the top two generation probabilities is equal to or greater than the threshold value of 90%. Therefore, it is possible to omit from the first image production parameters such as production probability indicating that there is no possibility of production of the target material, and it is possible to suppress useless provision of information to the user.
  • the second information acquisition unit 106 acquires second information indicating the composition formula and attributes of the known material, and the output unit outputs (i) the composition formula indicated by the second information of each of the one or more known materials. (ii) a map showing the relationship between the composition formula and the attribute indicated by the second information of each of the one or more known materials; or (iii) the second information of each of the one or more known materials A map showing the relationship of the displayed attributes is generated and output to the display unit 108 .
  • the map of (iii) is the map ma shown in FIG.
  • the second information is displayed on the map, so that the user can easily find the desired second information from the second information. be able to.
  • the user can easily select the desired second information and easily view the second image (that is, the reference image) corresponding to the second information.
  • the attributes of the known material indicated by the second information include the crystal structure of the known material, process conditions in the production process for producing the known material, characteristic values indicating the degree of characteristics of the known material, and the known material. and/or uses of known materials.
  • the process conditions are conditions used in the production process, such as temperature and time.
  • the user can know the attributes of known materials, and can easily guess information that will serve as a reference for the production of the target material based on those attributes.
  • the output unit identifies the types of generation processes that can be executed by the generation device owned by the user, and determines the generation parameters of the identified type of generation process among the generation parameters of each of the one or more types of generation processes.
  • the second information acquisition unit 106 acquires estimation information related to estimation of a generation device that can be executed by remaining generation processes other than the specified type of generation process among the one or more types of generation processes,
  • the output unit outputs an image showing the estimate information to the display unit 108 .
  • the user can easily request an estimate for a generator that the user does not own, according to the displayed estimate information.
  • the second information of each of the one or more known materials includes an estimated required time that is the time required to generate the known material
  • the output unit outputs the second information of each of the one or more known materials to the second information of the known material.
  • the pieces of information are arranged in order according to the estimated required time indicated by the two pieces of information, and a first image showing one or more pieces of second information arranged in the order is generated.
  • the estimated required time for each of one or more known materials is displayed, so the user can predict the time required to produce the target material based on the estimated required times.
  • the estimated required times are arranged and displayed in order, the user can easily grasp the maximum value, minimum value, or variance of the estimated required times, which is necessary for generating the target material. Time can be predicted better.
  • the second information of each of the one or more known materials also indicates a production process for producing the known material, any production process shortens the time required to produce the target material.
  • the user can easily determine whether the For example, even if the user intends to use the generation process with the largest generation parameter shown in the first image to generate the target material, if the estimated time required for the generation process is long, the next largest generation parameter may be generated. It can be determined that the process is used to produce the target material. In other words, the user can select the production process of the target material in consideration of the time required for production.
  • each component may be configured with dedicated hardware or realized by executing a software program suitable for each component.
  • Each component may be implemented by a program execution unit such as a CPU (Central Processing Unit) or processor reading and executing a software program recorded on a recording medium such as a hard disk or semiconductor memory.
  • a program execution unit such as a CPU (Central Processing Unit) or processor reading and executing a software program recorded on a recording medium such as a hard disk or semiconductor memory.
  • each component may be configured with dedicated hardware or realized by executing a software program suitable for each component.
  • Each component may be implemented by a program execution unit such as a CPU (Central Processing Unit) or processor reading and executing a software program recorded on a recording medium such as a hard disk or semiconductor memory. 5 to 7, 9, 11, 13, 16, 18, 23, and 23, the software programs for realizing the generation process display device and the database construction device of each of the above embodiments. 24 and the steps of the flow chart shown in FIG. 27 are executed by the computer.
  • a program execution unit such as a CPU (Central Processing Unit) or processor reading and executing a software program recorded on a recording medium such as a hard disk or semiconductor memory.
  • At least one of the above devices is a computer system specifically composed of a microprocessor, ROM (Read Only Memory), RAM (Random Access Memory), hard disk unit, display unit, keyboard, mouse, etc. be.
  • a computer program is stored in the RAM or hard disk unit.
  • At least one of the above devices achieves its functions by a microprocessor operating according to a computer program.
  • the computer program is constructed by combining a plurality of instruction codes indicating instructions to the computer in order to achieve a predetermined function.
  • a part or all of the components constituting the at least one device may be configured from one system LSI (Large Scale Integration).
  • a system LSI is an ultra-multifunctional LSI manufactured by integrating multiple components on a single chip. Specifically, it is a computer system that includes a microprocessor, ROM, RAM, etc. . A computer program is stored in the RAM. The system LSI achieves its functions by the microprocessor operating according to the computer program.
  • a part or all of the components that make up at least one of the devices described above may be composed of an IC card or a single module that is detachable from the device.
  • An IC card or module is a computer system composed of a microprocessor, ROM, RAM, and the like.
  • the IC card or module may include the super multifunctional LSI.
  • the IC card or module achieves its function by the microprocessor operating according to the computer program. This IC card or this module may have tamper resistance.
  • the present disclosure may be the method shown above. Moreover, it may be a computer program for realizing these methods by a computer, or it may be a digital signal composed of a computer program.
  • the present disclosure includes computer-readable recording media for computer programs or digital signals, such as flexible discs, hard disks, CD (Compact Disc)-ROM, DVD, DVD-ROM, DVD-RAM, BD (Blu-ray ( (registered trademark) Disc), semiconductor memory, etc. Alternatively, it may be a digital signal recorded on these recording media.
  • the present disclosure may transmit computer programs or digital signals via electric communication lines, wireless or wired communication lines, networks typified by the Internet, data broadcasting, and the like.
  • it may be implemented by another independent computer system by recording the program or digital signal on a recording medium and transferring it, or by transferring the program or digital signal via a network or the like.
  • the present disclosure has the effect of being able to appropriately support a user's search for a material production process, and can be used in a computer device or system for displaying information about the production process.
  • Second Generated Information 10 10a Generated Process Search System 100, 100a Generated Process Display Device (Generated Process Output Device) 101 first input unit 102 first information acquisition unit 103 composition descriptor generation unit 103a descriptor generation unit 104 firing learning unit 104a generation learning unit 105 probability derivation unit (derivation unit) 106 second information acquisition unit 107 image generation unit (output unit) 108 display unit 109 second input unit 110 input unit 200, 200a database construction device 202 article acquisition unit 203 region division unit 204 region descriptor generation unit 205 label assignment unit 206 combined information learning unit 206a generated information learning unit 207 label acquisition unit 208 Information extraction unit 301 First storage unit 302, 302a Second storage unit 401 Network d1 Target composition area d2 Known material area d3 Generation probability area d5 Quotation request area d6, d7 Check box d61 Pull-down d8 Threshold frame da1 Raw material area da

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Computing Systems (AREA)
  • Databases & Information Systems (AREA)
  • General Physics & Mathematics (AREA)
  • Physics & Mathematics (AREA)
  • Strategic Management (AREA)
  • Human Resources & Organizations (AREA)
  • Economics (AREA)
  • Data Mining & Analysis (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • General Engineering & Computer Science (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • General Business, Economics & Management (AREA)
  • Tourism & Hospitality (AREA)
  • Marketing (AREA)
  • General Health & Medical Sciences (AREA)
  • Primary Health Care (AREA)
  • Manufacturing & Machinery (AREA)
  • Health & Medical Sciences (AREA)
  • Development Economics (AREA)
  • Game Theory and Decision Science (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Operations Research (AREA)
  • Quality & Reliability (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
PCT/JP2022/006381 2021-02-22 2022-02-17 生成プロセス出力装置、生成プロセス出力方法、及びプログラム Ceased WO2022176945A1 (ja)

Priority Applications (4)

Application Number Priority Date Filing Date Title
JP2023500920A JPWO2022176945A1 (https=) 2021-02-22 2022-02-17
CN202280015567.9A CN116868220A (zh) 2021-02-22 2022-02-17 生成工艺输出装置、生成工艺输出方法及程序
EP22756263.4A EP4296933A4 (en) 2021-02-22 2022-02-17 PREPARATION PROCESS OUTPUT DEVICE, PREPARATION PROCESS OUTPUT METHOD, AND PROGRAM
US18/234,113 US12494271B2 (en) 2021-02-22 2023-08-15 Generation process output device, generation process output method, and recording medium

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
JP2021026720 2021-02-22
JP2021-026720 2021-02-22

Related Child Applications (1)

Application Number Title Priority Date Filing Date
US18/234,113 Continuation US12494271B2 (en) 2021-02-22 2023-08-15 Generation process output device, generation process output method, and recording medium

Publications (1)

Publication Number Publication Date
WO2022176945A1 true WO2022176945A1 (ja) 2022-08-25

Family

ID=82932277

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/JP2022/006381 Ceased WO2022176945A1 (ja) 2021-02-22 2022-02-17 生成プロセス出力装置、生成プロセス出力方法、及びプログラム

Country Status (5)

Country Link
US (1) US12494271B2 (https=)
EP (1) EP4296933A4 (https=)
JP (1) JPWO2022176945A1 (https=)
CN (1) CN116868220A (https=)
WO (1) WO2022176945A1 (https=)

Families Citing this family (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US12587274B2 (en) 2023-03-28 2026-03-24 Quantum Generative Materials Llc Satellite optimization management system based on natural language input and artificial intelligence
US12368503B2 (en) 2023-12-27 2025-07-22 Quantum Generative Materials Llc Intent-based satellite transmit management based on preexisting historical location and machine learning
US12603701B2 (en) 2023-12-27 2026-04-14 Quantum Generative Materials Llc Distributed satellite constellation management and control system
CN120595752B (zh) * 2025-08-06 2025-10-24 河北冀衡化学股份有限公司 生产设备的控制方法、系统、设备及可读存储介质
CN121029834A (zh) * 2025-10-31 2025-11-28 香港科技大学(广州) 用于原子级加工的工艺数据推送方法、装置、存储介质及设备

Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH1153425A (ja) 1997-08-05 1999-02-26 Sumitomo Chem Co Ltd 合成ルート設計システム及び合成ルート設計プログラムを記録した機械読み取り可能な記録媒体
JP2002163553A (ja) * 2000-11-28 2002-06-07 Toray Ind Inc 化合物商品案内システム、その案内方法、ならびに化合物商品の商取引システムおよびその商取引方法
JP2005157424A (ja) * 2003-11-20 2005-06-16 Sumitomo Chemical Co Ltd 有機化合物情報の処理システム、処理方法および処理プログラム
JP2009098924A (ja) * 2007-10-17 2009-05-07 Sumitomo Forestry Co Ltd 部材割付システム
JP2019179319A (ja) * 2018-03-30 2019-10-17 富士通株式会社 予測モデル作成装置、予測モデル作成方法および予測モデル作成プログラム
WO2020031671A1 (ja) * 2018-08-08 2020-02-13 パナソニックＩｐマネジメント株式会社 材料記述子生成方法、材料記述子生成装置、材料記述子生成プログラム、予測モデル構築方法、予測モデル構築装置及び予測モデル構築プログラム
JP2020038493A (ja) * 2018-09-04 2020-03-12 横浜ゴム株式会社 物性データ予測方法及び物性データ予測装置
JP2020077257A (ja) * 2018-11-08 2020-05-21 ジャパンモード株式会社 材料物性推定プログラム、材料生成機構推定プログラム
JP2020185573A (ja) * 2019-05-10 2020-11-19 オーエム金属工業株式会社 自動材料選択装置及び自動材料選択プログラム

Family Cites Families (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP5099511B2 (ja) * 2008-06-25 2012-12-19 国立大学法人山口大学 合成経路評価システムとその方法とそのプログラム
EP3746772A4 (en) * 2018-01-30 2021-11-10 Peter Madrid COMPUTED GENERATION OF CHEMICAL SYNTHESIS PATHS AND METHODS
US11735292B2 (en) * 2018-08-07 2023-08-22 International Business Machines Corporation Intelligent personalized chemical synthesis planning
WO2021124392A1 (ja) * 2019-12-16 2021-06-24 日本電信電話株式会社 材料開発支援装置、材料開発支援方法、および材料開発支援プログラム
US11610652B2 (en) * 2020-05-06 2023-03-21 Toyota Research Institute, Inc. Autonomous inorganic material synthesis machine
US20220359045A1 (en) * 2021-05-07 2022-11-10 International Business Machines Corporation Prediction of enzymatically catalyzed chemical reactions

Patent Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH1153425A (ja) 1997-08-05 1999-02-26 Sumitomo Chem Co Ltd 合成ルート設計システム及び合成ルート設計プログラムを記録した機械読み取り可能な記録媒体
JP2002163553A (ja) * 2000-11-28 2002-06-07 Toray Ind Inc 化合物商品案内システム、その案内方法、ならびに化合物商品の商取引システムおよびその商取引方法
JP2005157424A (ja) * 2003-11-20 2005-06-16 Sumitomo Chemical Co Ltd 有機化合物情報の処理システム、処理方法および処理プログラム
JP2009098924A (ja) * 2007-10-17 2009-05-07 Sumitomo Forestry Co Ltd 部材割付システム
JP2019179319A (ja) * 2018-03-30 2019-10-17 富士通株式会社 予測モデル作成装置、予測モデル作成方法および予測モデル作成プログラム
WO2020031671A1 (ja) * 2018-08-08 2020-02-13 パナソニックＩｐマネジメント株式会社 材料記述子生成方法、材料記述子生成装置、材料記述子生成プログラム、予測モデル構築方法、予測モデル構築装置及び予測モデル構築プログラム
JP2020038493A (ja) * 2018-09-04 2020-03-12 横浜ゴム株式会社 物性データ予測方法及び物性データ予測装置
JP2020077257A (ja) * 2018-11-08 2020-05-21 ジャパンモード株式会社 材料物性推定プログラム、材料生成機構推定プログラム
JP2020185573A (ja) * 2019-05-10 2020-11-19 オーエム金属工業株式会社 自動材料選択装置及び自動材料選択プログラム

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
See also references of EP4296933A4

Also Published As

Publication number Publication date
CN116868220A (zh) 2023-10-10
US12494271B2 (en) 2025-12-09
EP4296933A1 (en) 2023-12-27
US20230386617A1 (en) 2023-11-30
EP4296933A4 (en) 2024-08-21
JPWO2022176945A1 (https=) 2022-08-25

Similar Documents

Publication Publication Date Title
WO2022176945A1 (ja) 生成プロセス出力装置、生成プロセス出力方法、及びプログラム
Li et al. AI applications through the whole life cycle of material discovery
Xie et al. DeepVF: a deep learning-based hybrid framework for identifying virulence factors using the stacking strategy
Engel et al. Chemoinformatics: basic concepts and methods
Carbone et al. T-BAS: Tree-Based Alignment Selector toolkit for phylogenetic-based placement, alignment downloads and metadata visualization: an example with the Pezizomycotina tree of life
Lomize et al. OPM: orientations of proteins in membranes database
EP4027295A1 (en) Material property prediction device and material property prediction method
CN103838805B (zh) 医疗记录搜索方法和医疗记录搜索装置
Benafan et al. Shape memory materials database tool—a compendium of functional data for shape memory materials
WO2020203922A1 (ja) 結晶形予測装置、結晶形予測方法、ニューラルネットワークモデルの製造方法、及びプログラム
FR2822973A1 (fr) Procede, systeme et produit logiciel pour creer et visualiser un fichier graphique intelligent comprenant des informations relatives a des pieces
Mishra et al. Natural language processing and machine learning-based solution of cold start problem using collaborative filtering approach
JP2026002901A (ja) 情報処理装置、情報処理方法、及びプログラム
Geffers et al. Web-based digital gene expression atlases for the mouse
Zitouni et al. Integrating the opposition nelder–mead algorithm into the selection phase of the genetic algorithm for enhanced optimization
Rezaei et al. Immunoinformatics of Cancers: Practical machine learning approaches using R
US20040170949A1 (en) Method for organizing and depicting biological elements
US20250356225A1 (en) Answer generation method and system
Baird The impact of high‐throughput sequencing technology on speciation research: maintaining perspective
Whitelaw Visualising archival collections: The visible archive project
Santiago Galicia et al. Machine learning-driven prediction of heat transfer coefficients for pure refrigerants in diverse heat exchangers types
Maddison The rapidly changing landscape of insect phylogenetics
CN113010163B (zh) 材料测试表征和制备工艺数据库低代码构建方法和系统
Hu et al. Analyzing networks with VisANT
WO2023008449A1 (ja) 情報表示方法、情報表示装置およびプログラム

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 22756263

Country of ref document: EP

Kind code of ref document: A1

DPE1 Request for preliminary examination filed after expiration of 19th month from priority date (pct application filed from 20040101)
WWE Wipo information: entry into national phase

Ref document number: 2023500920

Country of ref document: JP

WWE Wipo information: entry into national phase

Ref document number: 202280015567.9

Country of ref document: CN

WWE Wipo information: entry into national phase

Ref document number: 2022756263

Country of ref document: EP

NENP Non-entry into the national phase

Ref country code: DE

ENP Entry into the national phase

Ref document number: 2022756263

Country of ref document: EP

Effective date: 20230922

WWW Wipo information: withdrawn in national office

Ref document number: 2022756263

Country of ref document: EP