JPWO2019049856A1 - Material specification information server, material selection support method and material selection support system - Google Patents

Abstract

Provided is a material specification information server and a material selection support method for supporting easy specification of appropriate specification items for various product applications. The material specification information server includes a condition decomposition unit 310 that generates a search term string from the material use conditions, a specification information database 320 that outputs a material property string from the search term string, and a predetermined specification item candidate string from the material property string. , A search word string has a plurality of pairs of registered words and their weights included in a registered word file, and the specification information database stores a search word string and a first mode correlation matrix. A plurality of first mode evaluation values for the search word string are calculated using the file and the plurality of first mode evaluation values and the characteristic-first mode correlation matrix file are used to calculate a plurality of characteristics for the search word string. A first evaluation value is calculated, and a material property sequence including a plurality of pairs of properties and each of the first evaluation values is generated. The candidate output unit outputs a second item from the material property sequence as a specification item candidate of the specification item candidate sequence. 1 High evaluation value To select a number of characteristics.

Description

  The present invention relates to a material specification information server, a material selection support method, and a material selection support system that support selection of a material used for a product.

  In order to improve the performance, increase the reliability, or reduce the price of industrial products, it is important to use appropriate materials. The types of materials used in industrial products are enormous, and the number is increasing year by year through research and development. Materials Electronic Commerce (hereinafter, referred to as "material EC") is a system by which a material consumer can select and purchase necessary materials from various types of materials through a communication line. As a material selection system for selecting an appropriate material from various materials via a communication line, there is a technique described in Patent Document 1.

  Patent Literature 1 discloses a consumer terminal that transmits a material selection request specifying a use, a material database that stores material information necessary for material selection for each material use, and an application for a material selection request specifying a use. A material selection system including a material selection mediation server for selecting an optimum material by searching for material information corresponding to the information from a material database and transmitting the selected material to a consumer terminal. In Patent Document 1, material information includes know-how based on material use experience, and know-how includes a past material damage case and a material processing problem case showing a problem in material processing in the past.

JP 2003-30248 A

  In Patent Literature 1, in order to input material selection conditions according to a use, a material selection mediation server acquires use information of a consumer, and displays a selection condition input screen corresponding to the use on a consumer terminal. It is said to be transmitted sequentially. Although the selection condition input screen is illustrated, there is no generalized description on how the material selection mediation server determines the material selection condition according to the use from the use information.

  Assuming that the material database stores the material selection condition for each application, the material selection condition may be determined from the application information by referring to the material database. Such a method is also practical if a material is selected only for a material for a specific use. However, in this method, it is necessary to prepare material selection conditions (hereinafter, referred to as “specification items”) corresponding to each increase in the use. In the case of the material EC that meets a wide range of consumers having various uses, it is difficult to determine the specification items for each such use one by one and create a database.

  Materials have many characteristics that can be specified as specification items. For example, mechanical properties alone include yield strength, tensile strength, total elongation, uniform elongation, and hardness. In addition to mechanical properties, there are thermal properties such as thermal conductivity, electrical properties such as electrical conductivity, and magnetic properties such as magnetic permeability. It is almost impossible to specify all these characteristics in the specification item uniformly. In addition, if characteristics that are unnecessary for the application are specified in the specification items, the optimal materials that should have been originally selected are eliminated by excessive specification items, or materials with excessive performance are selected and procurement costs increase. there is a possibility.

  Furthermore, when selecting materials to purchase, material consumers should specify specification items that do not leak out as long as they meet the intended use.However, material users should have sufficient knowledge to make optimal specifications. It is not always necessary to have it, and it is desired to have a function to support specification of specification items even in such a case.

  For example, in order to reduce the development cost or production cost of a product using the material, or to reduce defects in the use of the product, from the initial development stage, a later stage of product development such as actual machine prototype, or a product It is preferable that the specification items be completely specified in consideration of a defect that may occur during a later manufacturing stage or during use of the product. However, defects that occur during actual prototype production, manufacture, or use of a product often occur depending on the characteristics of the material depending on the processing applied to the material and the use environment and use conditions of a product using the material. For this reason, even if the user who selects the material is an engineer in the field, it is not easy to comprehensively grasp the response of the material to the processing, the use environment, and the use condition.

  In the present invention, for a variety of product applications, a user who selects a material is not necessarily familiar with the latter stage of product development, such as prototype production of an actual machine, or a manufacturing stage of a product, or of a product. An object of the present invention is to provide a material specification information server supported by a material selection support system and a material selection support method using the same so that specification items can be easily specified based on characteristics required in a use stage.

  In order to solve the above-mentioned problem, a material support information server according to an embodiment includes a registered word file for registering a plurality of registered words, and a registered word file for registering correlation values between a plurality of registered words and a plurality of first modes related to a product. A one-mode correlation matrix file, a property for registering correlation values between a plurality of properties relating to a material and a plurality of first modes—a data file storing a first-mode correlation matrix file, and a search term based on material usage conditions input from a terminal A condition decomposing unit that generates a sequence, a specification information database that outputs a material property sequence from the search word sequence, and a candidate output unit that outputs a predetermined specification item candidate sequence to the terminal from the material property sequence, Has a plurality of pairs of registered words and their weights included in the registered word file, and the specification information database uses the search word string and the first mode correlation matrix file to store a plurality of pairs of the search words. A mode evaluation value is calculated, a first evaluation value of a plurality of characteristics for a search word string is calculated using the evaluation values of the plurality of first modes and the characteristic-first mode correlation matrix file, and the plurality of characteristics are respectively calculated. The candidate output unit selects a predetermined number of characteristics having a high first evaluation value from the material characteristic sequence as a specification item candidate of the specification item candidate sequence.

  Other problems and novel features will be apparent from the description of this specification and the accompanying drawings.

  It is possible to provide a material specification information server and a material selection support method for easily specifying appropriate specification items for various product applications.

1 is a diagram illustrating a basic configuration of a material selection support system according to a first embodiment. FIG. 4 is a sequence diagram illustrating an operation in the material selection support system. It is a flowchart which shows the detail of a condition decomposition | disassembly process. FIG. 4 is a diagram illustrating a method for generating a search word string from a word string. It is a flowchart which shows the detail of specification information database search processing. It is an example of a damage mode correlation matrix D. It is an example of a characteristic-damage correlation matrix P. It is a flowchart which shows the detail of a priority order determination process. It is a data file stored in a material specification information server. It is an example of a screen of a material use condition input section and a specification item determination section. FIG. 4 is a sequence diagram illustrating an operation in the material selection support system. It is an example of a screen of a material use condition input section and a specification item determination section. It is an example of a damage / quality mode correlation matrix DQ. It is an example of a characteristic-damage / quality correlation matrix PR. FIG. 9 is a diagram illustrating a basic configuration of a material selection support system according to a second embodiment. FIG. 4 is a sequence diagram illustrating an operation in the material selection support system. It is an example of the screen of a material search part. It is an example of a screen of a material selection section. FIG. 4 is a sequence diagram illustrating an operation in the material selection support system. It is an example of a damage mode / material correlation matrix DU. It is an example of a screen of a material use condition input section and a specification item determination section. FIG. 14 is a diagram illustrating a basic configuration of a material selection support system according to a fourth embodiment. It is an example of sales material information. It is a flowchart which shows an example of a material ordering procedure. It is an example of an area fare table in the small-lot delivery service. It is an example of the charge table according to the size of goods. It is a data flow diagram in which a procurement information calculation part uses an estimation service via API. It is an example of a material procurement plan table PT. It is an example of a quote.

  Hereinafter, embodiments will be described with reference to the drawings.

  FIG. 1 illustrates a basic configuration of the material selection support system according to the first embodiment. The consumer terminal 100 is connected to the material specification information server 300 via the communication line 200. The consumer terminal 100 has a material use condition input unit 110 and a specification item determination unit 120. The material use condition input unit 110 and the specification item determination unit 120 are realized as a web application executed on a web browser of the terminal 100, but are not limited thereto, and may be implemented as a stand-alone application. The material specification information server 300 includes a condition decomposition unit 310, a specification information database system 320, and a candidate output unit 330. The condition decomposition unit 310, the specification information database system 320, and the candidate output unit 330 are realized as software executed by a general computer architecture. That is, the programs corresponding to the condition decomposition unit 310, the specification information database system 320, and the candidate output unit 330 are stored in the storage device of the material specification information server 300, and the processor of the material specification information server 300 reads out these programs. Execute. The material specification information server 300 may be realized on a single computer or on a cloud.

  FIG. 2 is a sequence diagram showing an operation in the material selection support system of FIG. When the customer inputs the material use conditions from the material use condition input unit 110 of the terminal 100 (S110), the input material use conditions are transmitted to the material specification information server 300 via the communication line 200.

  The “material use condition” includes at least the name of a product or a part manufactured by using the material (hereinafter, referred to as a “product” unless otherwise specified). In order to improve the accuracy of the specification item candidates output from the material specification information server 300, the “material use conditions” include, in addition to the name of the product, the environment in which the product is used and the material used to manufacture the product. It is desirable to include additional processing. Here, it is assumed that the present material selection support system handles a variety of materials widely, and the environment to be used is included in the environment as long as the contents affect the selection of materials. For example, when a product manufactured using a material is a part, the name of the product in which the part is used is also included.

  The condition decomposition unit 310 of the material specification information server 300 performs a condition decomposition process (S310) on the received material use conditions to generate one or more search terms. The material specification information server 300 searches the specification information database system 320 using the generated search word (S320), and obtains a correlation between a plurality of material properties and the search word. The candidate output unit 330 of the material specification information server 300 determines a priority as a specification item for a plurality of material characteristics based on the obtained correlation (S330), and determines a predetermined number of material characteristics having a high priority. , And transmits to the terminal 100 as specification item candidates that match the material use conditions input by the consumer.

  The consumer checks the specification item candidates displayed on the specification item determination unit 120 of the terminal 100 (S120), and judges whether or not it is necessary to change the material use condition input first (S121). When it is not necessary to change the material use condition, the specification item is determined (S122), and when the use condition is changed, the material use condition is input again (S110).

  FIG. 3A is a flowchart illustrating details of the condition decomposition processing (S310) performed by the condition decomposition unit 310. When the condition decomposition process is started (S311), the process enters a standby state and waits for input of a material use condition. When the material use condition is input, the material use condition is subjected to morphological analysis (S312) to be decomposed into one or more word strings. By the morphological analysis, the input sentence is decomposed into a row of the original form of the word registered in the dictionary file (the language of the dictionary file corresponds to the language of the material use condition input in the material selection support system). Identify the part of speech of each word. When “body of organic solvent pump” is input as the material use condition, morphological analysis is performed using a morphological analysis dictionary file stored in the material specification information server 300, and word strings (“organic solvent”, “pump”) "Body").

  After initializing the search word string, it is confirmed whether or not the word string output by the morphological analysis is registered in a registered word (registered key) file previously stored in the material specification information server 300 (S313). . If the word is a registered word, the output word is added to the search word string, and the similarity of the word is set to 1 (S314). A search word sequence (keyword sequence) is a set of pairs of a search word (keyword) and its weight (weight).

  If the word is not a registered word, the similarity between the word and the registered word is calculated (S315), and a registered word having a high similarity is added to the search word string (S316). It is checked whether the processing has been completed for all the words (S317), and if not, the process returns to step S313 to process the next word.

For calculating the similarity between words, for example, a cosine similarity known in the field of natural language processing can be used.
Sij = (Vi · Vj) / (| Vi || Vj |)
Here, Sij is the similarity between word i and word j, Vi (Vj) is the vector representing word i (word j), Vi · Vj is the inner product of the two vectors, | Vi | (| Vj | ) Is the absolute value of the vector Vi (vector Vj). The technique of expressing words as vectors is known in the field of natural language processing as distributed representation or word embedding. The material specification information server 300 has a sentence database file (corpus file) in order to represent words by vectors. As the corpus file, a database of sentences (static corpus) previously collected from engineering literature and the like, a sentence database (dynamic corpus) storing results of searching for the word on the Internet, or both can be used. .

  In order to calculate the similarity between the words output by the morphological analysis and the registered words, the words in the word string are expressed as vectors using a corpus file. On the other hand, the vector representation of the registered words based on the corpus file is stored in the registered word file together with the registered words.

  The process of generating a search word string will be described with reference to FIG. 3B, taking a word string (“organic solvent”, “pump”, “body”) as an example. Here, it is assumed that “organic solvent” and “body” are not registered in the registered word file, and “pump” is registered.

  Since “organic solvent” is not registered in the registered word file (S313), the similarity with the registered word is calculated in step S315. As a result, a registered word group 3002 similar to “organic solvent” is extracted. When the setting is such that a registered word having a similarity of not less than the threshold value 0.7 is added to the search word, the search word string derived from “organic solvent” is the search word string 3004-1. Note that a predetermined threshold may be set for the degree of similarity in order to determine whether or not to be added as a search word, or an adjustment may be made so that an appropriate number of registered words are added as search words.

  Since “pump” is registered in the registered word file (S313), the search word string derived from “pump” is the search word string 3004-2.

  Since “body” is not registered in the registered word file (S313), the similarity with the registered word is calculated in step S315. As a result, a registered word group 3003 similar to “body” is extracted. A registered word having a high degree of similarity is extracted, and a search word string derived from “body” is a search word string 3004-3.

  Thus, the search term does not always match the input word. Therefore, as an index indicating the closeness between a search word and a word, a similarity is set as a search word string. However, since the similarity is a value obtained for each word, the similarity is normalized (weight normalization) to express the proximity to the word string in the entire search word string as an index. Weight. Weight normalization is performed in two stages. In the first normalization, the similarity of a search word corresponding to each word is normalized. In the normalization method, for example, each similarity is divided by the root of the sum of squares of the similarity of the search word corresponding to each word. The search word string 3005 indicates a value obtained by normalizing the similarity for each word as described above. In the case of “organic solvent”, the similarity of three search terms is divided by the root of the sum of squares of similarities of search terms corresponding to “organic solvent” (1.39).

  The second normalization is normalization for all search words. For example, the similarity normalized by the number of words in the word string is divided. The search word string 3006 indicates the value of the weight thus obtained. In this example, the normalized similarity is divided by the number 3 of words.

In step S318, the normalized search word string is formed into {{search word 1, weight 1},...} And output to the specification information database system 320. Therefore, a search word string shaped into a pair of a search word and its weight as follows:
｛{Alcohol, 0.215}, {oil, 0.191}, {volatile, 0.168}, {pump, 0.333}, {fuselage, 0.249}, {housing, 0.221}｝
Is output.

  After the step S318, the condition decomposition processing (S310) shifts to a standby state. Alternatively, the condition disassembly processing is terminated according to a management request of the server 300 (S319).

  FIG. 4A is a flowchart showing details of the specification information database search process (S320). When the specification information database search process is started (S321), the process enters a standby state and waits for input of a search word string. When the search word string is input, the search word string is stored in the memory as pre-search processing, and the damage correlation accumulation vector is initialized to 0 (S322). Details of the damage-correlation accumulation vector will be described later.

  After the search pre-processing, a damage mode correlation matrix D is searched for in the search word of the search word string (S323). FIG. 4B shows an example of the damage mode correlation matrix D. The damage mode correlation matrix D is a matrix containing correlation values between M registered words (registered keys) and N damage modes. The matrix value Dij is a correlation value between the registered word i (1 ≦ i ≦ M) and the damage mode j (1 ≦ j ≦ N), and takes a value from 0 to 1. A correlation value of 0 indicates that the damage mode j can be ignored for the registered word i, while a correlation value of 1 indicates that the damage mode j must be considered for the registered word i.

  The damage mode categorizes damages that occur during product development, product manufacture, and product use using the material, such as rupture, deformation, corrosion, wear, erosion, weld cracking, and fatigue fracture. and so on.

  The damage mode correlation matrix D can be created using a text mining technique. In the case of creating using text mining, for example, a co-occurrence probability that a registered word i and a damage mode j appear together in a plurality of prepared sentences may be used as Dij. it can. An expert in the engineering field may create or edit the damage mode correlation matrix D. For example, an expert in the engineering field may first create a damage mode correlation matrix D, and then extend the matrix D by text mining. Conversely, an expert can correct the damage mode correlation matrix D created by text mining.

  In the search processing (S323), a registered word that matches the search word is searched for from the damage mode correlation matrix D, and the row data of the damage mode correlation matrix D corresponding to the registered word (hereinafter, damage-correlation vector) Is the process that returns For example, when the a-th search word Ka matches the i-th registered word Ri, {Di1, Di2,..., DiN} is returned as the damage correlation vector corresponding to Ka. Note that the search word is guaranteed to be a registered word by executing the process described with reference to FIG. 3A.

  The addition process (S324) multiplies the damage correlation vector {Di1, Di2,..., DiN} corresponding to the search word Ka output from the search process (S323) by the weight Wa of the search word Ka to obtain the damage correlation cumulative vector. Add to A. The damage correlation accumulation vector A is a vector having the same number of elements as the damage correlation vector, and has been initialized to 0 in the preprocessing (S322).

  Next, it is confirmed whether or not the processing has been completed for all the search terms (S325). If not, the process returns to step S323 to process the next search term. In this way, the processing of steps S323 and S324 is applied to all search words in the search word string. When the search word string ends, the damage correlation accumulation vector A is output. The output damage correlation cumulative vector A = {A1, A2,..., AN} is a vector in which the necessity of considering each damage mode is numerically evaluated based on the entire search word string.

  In the multiplication process (S326), the property-damage correlation matrix P is multiplied by the damage correlation accumulation vector A. FIG. 4C shows an example of the property-damage correlation matrix P. The property-damage correlation matrix P is a matrix of Q rows and N columns containing correlation values between properties of Q materials and N damage modes. The matrix value Pkj is a correlation value between the characteristic k (1 ≦ k ≦ Q) and the damage mode j, and takes a value from 0 to 1. A correlation value of 0 indicates that the damage mode j is not related to the characteristic k, and a correlation value of 1 indicates that the damage mode j is strongly influenced by the characteristic k.

  The property is a property of a material, for example, chemical composition ratio, density, yield strength, tensile strength, hardness, elongation, carbon equivalent, carbon resistance, pitting resistance equivalent number, PREN ), Critical crevice corrosion temperature, melting temperature, solidifying temperature, glass transition temperature, thermal conductivity, specific heat, electrical conductivity, magnetic permeability, price , Price volatility, and the like.

  The property-damage correlation matrix P, like the damage mode correlation matrix D, can be created by machine learning using text mining, by an expert in the field of engineering, or both.

The multiplication of the property-damage correlation matrix P and the damage correlation accumulated vector A is performed by a general matrix-vector multiplication process in linear algebra. That is, the element value Tk of the material property vector (property vector) T = {T1, T2,..., TQ} obtained by the multiplication is calculated by the following equation.
Tk = Σj (Pkj × Aj), (j = 1, 2,..., N)
In the material property sequence output process (S327), a material property sequence including a pair of a material property corresponding to each element value of the material property vector Tk is created and output to the candidate output unit 330. The material property column is
{Characteristic 1, T1}, {Characteristic 2, T2}, ..., {Characteristic Q, TQ}
It is represented as It can be evaluated that the larger the element value Tk is, the more important the characteristic is.

  The specification information database search process (S320) shifts to a standby state after step S327. Alternatively, the processing can be terminated (S328) by a request on the management of the server 300 or the like.

FIG. 5 is a flowchart illustrating details of the priority order determination process (S330) executed by the candidate output unit 330. When the priority order determination process is started (S331), the process enters a standby state, and waits for input of a material property sequence. When a material property column is input, the elements are sorted in descending order of their element values Tk (S332). Next, the B items from the head of the material property sequence arranged in descending order are output as specification item candidate sequences (S333). The specification item candidate column is
{Specification item candidate 1, U1}, {Specification item candidate 2, U2}, ..., {Specification item candidate B, UB}
It is represented as

  The specification item candidate sequence is data obtained by arranging the material characteristic sequence in descending order of the element value Tk and then extracting B items from the beginning, so that U1 ≧ U2 ≧... ≧ UB. The output specification item candidate string is transmitted by the server 300 to the terminal 100 on the consumer side via the communication line 200.

  After the step S333, the priority order determination process (S330) shifts to a standby state. Alternatively, the processing can be terminated (S334) by a request on the management of the server 300 or the like.

  FIG. 6 shows a data file stored in the material specification information server 300. The material specification information server 300 includes a dictionary file 601 for morphological analysis used in S312 processing, a registered word file 602 used in S313 and S315 processing, a corpus file 603 used in S315 processing, and S323 and S324 processing. A damage mode correlation matrix D file 604 to be used and a characteristic-damage correlation matrix P file 605 to be used in the S326 process are stored.

  FIG. 7 is an example of a screen of the material use condition input unit 110 and the specification item determination unit 120 in the consumer terminal 100. In the example of FIG. 7, the material use condition input unit 110 and the specification item determination unit 120 are displayed on the upper and lower rows of one screen. Various modifications are possible, such as changing the arrangement of the screens and displaying the material use condition input unit 110 and the specification item determination unit 120 on different screens.

  The screen 701 of the material use condition input unit 110 shown in the upper part of FIG. 7 will be described. On the left side are menus for selecting the type of material use conditions. As described above, since the material use condition includes the product as an essential, the type of the first line 702 of the material use condition is fixed to the product. The second and subsequent lines (703, 704) can be selected from the use environment, processing method and the like. On the right side of the type menu is a text box 705 for inputting a search word. In addition to entering free sentences, you can also enter words as logical expressions combining operators (AND, OR, NOT, etc.) with parentheses. On the right side of the text box 705 is a menu of logical operators 706 that combine the conditions of multiple lines. The screen 701 shows an example of three lines. However, the present invention is not limited to this. Alternatively, a button for adding the number of lines may be arranged so that a condition line can be added as required by the customer.

  When the customer clicks a search button 707 at the upper right of the screen, the material use condition input unit 110 first checks whether a text has been entered in the search text box of the required product 702. If a text is input, the material use condition input unit 110 transmits the input contents of each search text box to the material specification information server 300. If no text has been entered in the product search text box, a message screen or voice or other signal prompting entry is output to the consumer.

The process in which the material specification information server 300 receives the search text, that is, the input of the usage condition of the material and outputs the candidate of the specification item is as described above. The processing can also be performed as a simple extension of the described processing. For the usage conditions of the plurality of rows 702 to 704, the input content 705 and the logical operator 706 for each row are transmitted to the specification information server 300 as usage conditions. The specification information server 300 performs the condition decomposition process of FIG. 3A on each line of the usage condition of a plurality of lines, and creates a search word string for each line. Next, the processing up to the specification information database search processing (S325) in FIG. 4A is performed for each search word string corresponding to each row, and a damaged interphase cumulative vector A for each row is created. Next, an operation between the damaged inter-phase accumulated vectors A is performed in accordance with the operator used for connecting the rows. For example, a combined damage interphase accumulated vector Ac obtained by combining the damage interphase accumulated vectors A and A ′ corresponding to two rows linked by AND by an element unit is obtained.
If A = {A1, A2,..., AN} and A ′ = {A′1, A′2,.
Ac = {Ac1, Ac2,..., AcN} = {A1 × A′1, A2 × A′2,.
Becomes

  The method of calculating vectors for obtaining the joint damage inter-phase cumulative vector Ac corresponds to the logical operator 706 used for connecting the rows. For example, vectors are added to two rows joined by OR, and vectors are subtracted to two rows joined by NOT. As described above, for the usage condition of a plurality of rows, the operation corresponding to the joint operator is performed on the corresponding damaged interphase accumulated vector, and one combined damaged interphase accumulated vector Ac is obtained in step S325 of FIG. 4A. Then, the processing after step S326 is performed using the combined damage interphase accumulated vector Ac.

  The screen 711 of the specification item determination unit 120 shown in the lower part of FIG. 7 will be described. Boxes (“item 1” to “item B”) 712 display specification item candidates 1 to B in the specification item candidate column received from the material specification information server 300. These boxes switch between a selected state and a non-selected state by clicking. For example, the initial state of each box is set to a non-selected state, and once a box is clicked, it becomes a selected state. Clicking the selected box again returns it to the non-selected state. .., And B are displayed in order from the material property having the highest evaluation value with respect to the input material use condition, so that the customer can check and select the item contents in order from item 1. . To assist the customer in confirming the item contents, detailed information on the item contents may be displayed in a separate window.

  When the confirmation and selection of the specification items are completed and the review is not necessary, the consumer clicks the determination button 713 to determine the specification items. If the consumer determines that the review is necessary, the input content of the material use condition input unit 110 is reviewed and searched again.

  With the above configuration and operation, by presenting the candidate of the specification item to the consumer using the material utilization condition input by the consumer, it is possible to assist the consumer to specify the specification item suitable for the material utilization condition. . Using the condition decomposition processing (S310) in the specification information server 300, the search text input by the consumer at the terminal 100 is converted into a search word string composed of registered words, and registered in advance in order to search the specification information database 320. It is possible to support the specification of the specification item even if the material use condition has not been set. This is a preferable feature for the material EC corresponding to a wide range of consumers having various product applications.

  Further, the specification information database system 320 has a damage mode correlation matrix D and a characteristic-damage correlation matrix P, and a damage information correlation vector A is calculated from a search word string in a specification information database search process (S320), and the damage correlation is calculated. A material property sequence is calculated from the accumulated vector A, and a specification item candidate sequence is created from the material property sequence.

  The material selection system disclosed in Patent Document 1 also includes past material damage cases and material processing problem cases as material information. However, the customer only displays the processing precautions and the damage case number for each material on the search result screen (the optimum material selection screen in Patent Document 1), and the specification items to be designated in order to prevent the damage. Is not presented to the consumer. For this reason, the consumer has to read the details of the damage case corresponding to the case number for each material, understand the contents and the physical phenomena, and reflect it on the material specification items by himself. This requires a high level of professional knowledge from the customer, and if the number of damage cases obtained by searching is large, work to read the details of the damage cases, understand the physical phenomena, and reflect them in the specification items Is a burden for consumers.

  On the other hand, in the present embodiment, the damage mode of the product to be noted in the material use condition of the customer is automatically incorporated into the specification item candidate. Required items can be easily specified.

FIG. 8A shows a sequence diagram as a modification of the first embodiment. The difference from the sequence diagram of FIG. 2 is that information transmitted from the specification information server 300 to the customer terminal 100 includes a damage mode sequence. The damage mode sequence is composed of the N damage modes stored in the damage mode correlation matrix D file 604 and the damage correlation cumulative vector A = {A1, A2,..., AN} output from the process S325 in FIG. Damage phase accumulated vector Ac = {Ac1, Ac2,..., AcN}), and has, for example, the following format.
{Damage mode 1, A1}, {Damage mode 2, A2}, ..., {Damage mode N, AN}}
The damage mode sequence can be created in, for example, the process S326 in FIG. 4A and output to the priority determination process (the process S330 in FIG. 8A) together with the material property sequence in the process S327. In the priority order determination process (see FIG. 5), the damage mode sequence may be arranged in descending order of the A value in the process S332, and a predetermined number of the damage mode sequences may be output to the consumer terminal 100 from the beginning in the process S333. This is because the larger the value of A is, the more important the damage mode is.

  FIG. 8B shows an example of a screen of the consumer terminal 100. The considered damage mode 802 is displayed on the screen 801 of the specification item determination unit 120. Thus, by presenting the information on the damage mode to the customer as the basis for selecting the specification item candidate, it is possible to assist the customer in specifying a more appropriate specification item.

  This is particularly advantageous for the material EC in which the consumer inputs only limited information. Since the material EC is a third party to the customer, the material usage conditions that the customer inputs to the material EC may be incomplete. By displaying the specification item candidate and the damage mode together on the screen of the specification item determination unit 120, the customer evaluates the specification item candidate and the damage mode using information that only the customer who has not entered the material EC knows. Thus, it is possible to determine the specification item or change the use condition.

  Next, a description will be given of a modification in which a quality mode is considered in addition to the damage mode in the search processing of the specification information database. Therefore, the damage / quality mode correlation matrix DQ shown in FIG. 9A is used instead of the damage mode correlation matrix D shown in FIG. 4B. Also, instead of the property-damage correlation matrix P shown in FIG. 4C, a property-damage / quality correlation matrix PR shown in FIG. 9B is used.

  The damage / quality mode correlation matrix DQ (FIG. 9A) is a matrix containing correlation values between M registered words, N damage modes, and K quality modes. The property-damage / quality correlation matrix PR (FIG. 9B) is a matrix that stores correlation values between the properties of Q materials, N damage modes, and K quality modes. Each of the correlation values takes a value from 0 to 1. The quality mode categorizes the quality required for a product, and includes, for example, light weight, small size, antibacterial, heat insulation, heat resistance, gloss, and transparency.

  In this modification, a damage / quality mode correlation matrix DQ file is stored in the specification information server 300 instead of the damage mode correlation matrix D file 604, and the characteristic-damage / quality is replaced with the characteristic-damage correlation matrix P file 605. The correlation matrix PR files are stored (see FIG. 6), and the contents of the processing are executed by replacing these files in FIG. 4A. First, in the pre-search process (S322) of FIG. 4A, the damage / quality correlation accumulation vector of (N + K) length is initialized to zero. Next, in the search process (S323), the damage / quality mode correlation matrix DQ is searched using the search word to obtain an (N + K) -length damage / quality correlation vector. Next, in processing S324, the damage / quality correlation vector is multiplied by the weight of the search word, and then added to the damage / quality correlation cumulative vector. Steps S323 and S324 are repeated with changing the search term until the search term string ends (S325). After the search for all the search terms is completed, the damage / quality correlation accumulated vector of (N + K) length is decomposed into the damaged inter-phase accumulated vector of length N and the quality inter-phase accumulated vector of length K.

  In processing S326, the damaged inter-phase accumulated vector and the quality inter-phase accumulated vector are respectively converted into a Q row × N column sub-matrix 903 corresponding to the damaged mode of the characteristic-damage / quality correlation matrix PR shown in FIG. 9B and a quality mode. Multiply the corresponding Q row × K column sub-matrix 904 to obtain the damage related material property vector TD = {TD1, TD2,..., TDQ} and the quality related material property vector TR = {TR1, TR2,. obtain. Since the material property vectors TD and TR have the same length and format as the material property vector T described with reference to FIG. 4A, the processes after the process S327 may be the same except that only two output vectors are used. Corresponding to the two output vectors, the specification item candidate presented from the damage-related material characteristics and the specification item candidate presented from the quality-related material characteristics are separately displayed on the screen of the specification item determination unit 120. . For example, in the specification item selection section shown in FIG. 7 or 8B, the line of the damage-related specification item and the line of the quality-related specification item may be displayed separately.

  9A and 9B show an example in which the correlation value for the damage mode and the correlation value for the quality mode are integrated into one matrix file, but the damage / quality mode correlation matrix DQ is represented by a partial matrix. 901 and a quality mode correlation matrix corresponding to the submatrix 902 may be stored as two matrix files. Similarly, the property-damage / quality correlation matrix PR may be stored as two matrix files of a property-damage correlation matrix corresponding to the sub-matrix 903 and a property-quality correlation matrix corresponding to the sub-matrix 904. Equivalent processing is possible regardless of the file format.

  8A and 8B, it is described that the considered damage mode string is displayed on the customer terminal 100. Similarly, the considered damage mode and the quality mode are displayed on the customer terminal 100 in the specification items. It can also be displayed on the screen of the determination unit 120. Thereby, not only avoidance of damage to the product but also specification item candidates in consideration of the quality of the product can be presented to the consumer, so that it is possible to assist the consumer in selecting a more appropriate specification item. For example, when “automobile shaft” is input as the material use condition, for example, “lightweight” having a high correlation with “automobile” is selected as the quality mode, and is a material characteristic having a high correlation with “lightweight”. It can be expected that “density” is added to the specification item candidates. In this case, the consumer can select not only specification items that are closely related to damage such as strength, but also specification items that are related to added value such as lightness.

  FIG. 10 illustrates a basic configuration of the material selection support system according to the second embodiment. The second embodiment is characterized in that, in addition to the determination of the specification items, the process includes selection of materials using the specification items. A material information server 400 is connected to the communication line 200 in addition to the consumer terminal 100 and the material specification information server 300. The material information server 400 has a material information database 410. The material information database 410 stores values for a plurality of materials for each material property. The material properties in the material information database 410 match the Q material properties of the property-damage correlation matrix P (or property-damage / quality correlation matrix PR) stored in the specification information database system 320.

  Although the material specification information server 300 and the material information server 400 are separately described here for the purpose of clarifying the configuration, the material specification information server 300 also serves as the material information server 400, or the material information server 400 The information server 300 may also be used.

  Further, the consumer terminal 100 includes a material search unit 130 and a material selection unit 140 in addition to the material use condition input unit 110 and the specification item determination unit 120. The material search unit 130 and the material selection unit 140 are typically web applications used on the web browser of the consumer terminal 100, but may be standalone applications.

  FIG. 11 is a sequence diagram showing an operation in the material selection support system of FIG. The process from the process S110 in which the customer inputs the material use condition from the material use condition input unit 110 of the terminal 100 to the process S122 in which the customer determines the specification item is the same as the first embodiment.

  After determining the specification item (S122), the consumer inputs a specification value for the determined specification item to the material search unit 130 of the terminal 100. At this time, the specification item determined by the specification item determination unit 120 is displayed on the material search unit 130. When the consumer starts searching for a material (S131), a search condition including the specification item and the specification value is transmitted to the material information server 400. The search start is typically performed by clicking a search button on the screen, but is not limited thereto, and various methods such as voice and action (gesture) are possible.

  Upon receiving the search condition, the material information server 400 searches the material information database 410 using the search condition (S410), and extracts information on the material satisfying the search condition. Searching a database using a plurality of search conditions is possible using a known technique. A search result satisfying the search condition is transmitted to the consumer terminal 100.

  The search result received by the terminal 100 is displayed on the material search unit 130, and the consumer checks the material information obtained as the search result (S132) and determines whether the specification value needs to be changed (S133). If it is determined that the specification value needs to be changed, the consumer changes the specification value input to the material search unit 130 and starts searching (S131). If it is determined that the specification value does not need to be changed, the consumer moves to the material selection unit 140 and determines a purchase material (S141).

  FIG. 12 is an example of a screen of the material search unit 130. In this screen example, the specification item 1201 determined by the specification item determination unit 120 is displayed at the lower right of the screen, and a lower limit value 1202 and / or an upper limit value 1203 can be input for each item. When the consumer clicks the search button 1204, the specification items and specification values are stored in the memory of the terminal 100 as search conditions, and the search conditions are transmitted to the material information server 400 as described above.

  After the search result received from the material information server 400 is stored in the memory of the terminal 100, it is displayed on a part 1210 (left half) of the screen of the material search unit 130. A graph 1211 at the upper left of the screen displays the number of materials that satisfy the search condition. In reality, it is rare that a material search can be completed by a single search, and it is usual to perform a search while changing a specification value to change a material narrowing range. The screen shown in FIG. 12 allows the user to easily check this search history. In the graph 1211, the material search unit 130 assigns a serial number to the result saved by the consumer and displays the result with a specific mark, and does not assign a serial number to the unsaved result so that the result can be distinguished from the saved result. Will be displayed with a different mark. The consumer can click the save button 1212 to save a set of search conditions and search results. In addition, material information is displayed in a part 1213 (lower left of the screen) of the screen of the material search unit 130.

  The consumer checks the number of materials satisfying the search condition and the displayed information on the materials to determine whether or not the specification needs to be changed. If the customer determines that the specification change is unnecessary, the user moves to the material selection unit 140. In the screen example of FIG. 12, the customer inputs a result number for which detailed information is to be displayed in a result number input box 1214 at the upper right of the screen, and clicks a detail button 1215.

  FIG. 13 is an example of a screen of the material selection unit 140 according to the present embodiment. On the screen, “Result number and number of corresponding materials” 1301 being displayed is displayed. In the specification item 1302 and the specification value 1303, the search condition stored as the result number is displayed. In each column of the material 1304, material information of the search result stored as the result number is displayed for each specification item. A check box 1305 for selecting a material to be purchased is provided near the material name.

  The consumer can determine the material to be purchased by examining the specification values of the material satisfying the search condition from the screen of the material selection unit 140. If there is a material to be purchased, the consumer checks the check box 1305 of the material, and then clicks a decision button 1306 to determine the material to be purchased. If there is no material to purchase, the consumer clicks the return button 1307 and moves to the material search unit 130.

  In the second embodiment, after the specification item is determined, it is possible to easily select the purchased material using the specification item.

  In the third embodiment, input of specification values for material selection is facilitated by presenting reference material candidates. The basic configuration of the material selection support system of the third embodiment is the same as that of FIG. FIG. 14 is a sequence diagram illustrating an operation in the material selection support system according to the third embodiment. The process from the process S110 in which the consumer inputs the material use condition from the material use condition input unit 110 of the terminal 100 to the condition decomposition process (S310) in the specification information server 300 is the same as that of the first embodiment. In the specification information database search processing (S320), the specification information database is searched using the search word string output in the condition decomposition processing (S310), and the reference material string is added in addition to the material characteristic string described in the first embodiment. Also create Therefore, in the third embodiment, the damage mode / material correlation matrix DU shown in FIG. 15 is used instead of the damage mode correlation matrix D shown in FIG. 4B.

  The damage mode / material correlation matrix DU indicates a matrix containing correlation values between M registered words, N damage modes, and G materials. A sub-matrix 1501 of M rows and G columns of the damage mode / material correlation matrix DU is a matrix containing correlation values between M registered words and G specific materials. The value of the matrix is a correlation value between the registered word and the material, and takes a value from 0 to 1. A correlation value of 0 indicates that the registered word and the material are irrelevant, while a correlation value of 1 indicates that the registered word and the material have a deep relationship (for example, a frequently used material). In the third embodiment, it is possible to present a specific material name having a close relationship, thereby supporting a user's selection of a material. In the third embodiment, a damage mode / material correlation matrix DU file is stored in the specification information server 300 instead of the damage mode correlation matrix D file 604 (see FIG. 6). The contents of the processing are executed by replacing this file in FIG. 4A. The material correlation matrix corresponding to the sub-matrix 1501 may be stored as a file separate from the damage mode correlation matrix D. Equivalent processing can be performed regardless of the file format.

  In the search pre-processing (S322) of FIG. 4A, the damage / material correlation accumulation vector of (N + G) length is initialized to zero. Here, N is the number of damage modes, and G is the number of materials. Next, a damage mode / material correlation matrix DU is searched in a search process (S323) to obtain a damage / material correlation vector of (N + G) length. Next, in processing S324, the damage / material correlation vector is multiplied by the weight of the search word, and then added to the damage / material correlation cumulative vector. Steps S323 and S324 are repeated with changing the search term until the search term string ends (S325). After the search for all the search words is completed, in step S326, the damage / material correlation cumulative vector having a length N corresponding to the damage mode and the material correlation cumulative vector U having a length G corresponding to the material are determined. = {U1, U2,..., UG}. The process S326 outputs the material correlation accumulation vector U as it is to the process S327. In the process S326, similarly to the first embodiment, the damage correlation accumulation vector is multiplied by the property-damage correlation matrix P, and the multiplication result is output to the process S327 as a material property vector T.

The process S327 creates a reference material sequence in the same manner as the material characteristic sequence. The reference material column is
{{Material 1, U1}, {Material 2, U2}, ..., {Material G, UG}}
It is represented as The higher the element value Ul, the higher the relevance, for example, the more frequently used material can be evaluated.

  The reference material sequence is output to the priority determination process (S330) together with the material characteristic sequence. The priority order determination process (S330) creates a reference material candidate sequence in the same manner as the process of creating a specification item candidate sequence from a material property sequence. That is, after arranging the reference material sequence in descending order of the element value Ul, E is extracted from the head and transmitted to the terminal 100 as a reference material candidate sequence.

  The reference material candidate sequence is displayed on the specification item determination unit 120 of the terminal 100 together with the specification item candidate sequence. FIG. 16 is an example of a screen of the specification item determination unit 120 according to the present embodiment. The material 1602 of the reference material candidate column is displayed together with the specification item 1601 of the specification item candidate column on a part of the screen, and the consumer can select the reference material by clicking. Using this screen, the consumer confirms the reference materials together with the specification items (S120), and determines each of them (S122).

  When the specification item and the reference material are determined, the terminal 100 transmits the specification item and the reference material to the material information server 400. The material information server 400 performs the above-described material information database search processing (S410), extracts information on the reference material, and transmits the information to the terminal 100. The reference material information is displayed on the material search unit 130 of the terminal 100 together with the corresponding specification items. When a customer inputs a specification value for a specification item, the customer can input the value with reference to the reference material in the specification item. Alternatively, the material search unit 130 may have a means for inputting the specification value in a ratio with the value of the reference material in the specification item. This has the advantage that the consumer can easily set the specification value for the selected specification item.

  This is particularly advantageous when setting specification values for specification items that have little experience or expertise for the consumer. Since the specification value greatly changes depending on the type and unit of the specification item, the consumer needs to examine the related information to determine a proper value in order to input the specification value of the specification item with little experience or expertise. In the present embodiment, the user uses the material having a high correlation with the input material use condition as a reference material and displays the characteristic value of the reference material for each specification item in the material search unit 130, so that the customer can investigate the related information again. The specification values can be input without burdening the user.

  FIG. 17 illustrates a basic configuration of a material selection support system according to the fourth embodiment. The fourth embodiment is characterized in that, after the purchase material is determined in the second embodiment or the third embodiment, a procedure is performed until the purchase material is ordered from a material maker.

  The material selection support system includes a customer terminal 100, a material specification information server 300, a material information server 400, a material maker terminal 2000, and a distributor terminal 2500, which are connected to the communication line 200.

  The material information server 400 has a material information database 410, a procurement information database 420, and a procurement information calculation unit 430.

  The material information database 410 stores values for a plurality of materials for each material property. The material properties in the material information database 410 match the Q material properties of the property-damage correlation matrix P (or property-damage / quality correlation matrix PR) stored in the specification information database system 320.

  The procurement information database 420 is used for procurement such as a material maker that provides the material, sales material information of the material sold by the material maker, estimation of an order for the material itself, and transaction performance information (transaction date, etc.) with the material maker. Necessary procurement information is stored. The sales material information is data representing characteristics, specifications, performance, quality, and the like of various materials that can be provided by the material maker, as shown in FIG. 18, created based on information provided from a plurality of material manufacturers.

  FIG. 18 is an example of sales material information stored in the procurement information database 420, and shows an example of material characteristics relating to “SUS304 sheet material”. Thus, in the procurement information database 420, the sales material information is stored as a record for each material. The record for each material records the name of the maker providing the material, the place of manufacture (storage) of the material, the material name (model number), and the material characteristics. The material properties are further divided into chemical compositions (for example, Cr component ratio, Mo component ratio, etc.) constituting the material, and physical properties of the material (for example, density (specific gravity), hardness, etc.).

  In addition, each record for each material is not limited to the above information, but also includes identification information for identifying a company such as a period required for shipment, a delivery amount, a unit price, and a provider (eg, a company identification code, a company identification code). Number, etc.). The material characteristics recorded in each record include environmental characteristics (heat resistance, cold resistance, corrosion resistance, etc.), thermal conductivity, electrical conductivity, dimensional stability (linear expansion coefficient), aging characteristics, magnetization characteristics, and the like. It may be included. Further, the number of actual transactions performed may be recorded in each record.

  The sales material information may be distributed as paper documents such as catalogs and spec sheets, or may be distributed as electronic documents. The administrator of the material information server 400 may input, for example, information distributed, disclosed, and made public to the procurement information database 420 from each material maker. Further, the material specification information server 300 or the material information server 400 may acquire information from a website of each material maker using a computer program such as a search engine and store the acquired information in the procurement information database 420.

  The material maker terminal 2000 is a terminal installed for each material maker. The material maker terminal 2000 is a PC, a server, or the like used by the staff, employees, or the like of the company that is the material maker. The material maker receives a material procurement instruction via the material maker terminal 2000. In the example of FIG. 17, the material maker M1 as a procurement candidate has a material maker terminal 2000a, the material maker M2 has a material maker terminal 2000b, and the material maker M3 has a material maker terminal 2000c.

  The distributor terminal 2500 is a terminal installed for each distributor. The logistics company provides material transportation from a material manufacturer to a material receiving base specified by a customer. The logistics company manages the transportation of materials, the status of delivery contracts, the operation status of vehicles related to transportation, the completion of delivery, and the like. The distributor terminal 2500 is a PC, a WS, a server, or the like used by staff, employees, and the like of a business that is a distributor. The distributor accepts the delivery arrangement from the designated source to the destination via the distributor terminal 2500. Further, the distributor terminal 2500 can be connected via the network 200 to a portable terminal such as a smartphone or a vehicle-mounted terminal carried by the driver of the transport vehicle. In the example of FIG. 17, the distributor L1 that is a candidate for distribution has a distributor terminal 2500a, and the distributor L2 has a distributor terminal 2500b.

  Next, a material ordering procedure in this embodiment will be described. FIG. 19 is a flowchart illustrating an example of a material ordering procedure provided by the material information server 400. More specifically, the material to be purchased is determined according to the flowchart of FIG. 11 or FIG. 14 (S141), and the procedure from requesting a quote to a material maker providing the material to ordering the material is shown. Things.

  When the material to be purchased is determined (S141), the consumer inputs from the consumer terminal 100 whether to place an order for this material (S150). If no order is placed (N in S150), the process ends (S156). On the other hand, when ordering (Y in S150), the material to be ordered (material to be purchased) is selected (S151). When there are a plurality of materials determined to be purchased, all the materials may be ordered, or only some of the materials may be ordered. Subsequently, the consumer inputs the delivery destination of the ordered material from the consumer terminal 100 (S152). The delivery destination is not limited to the place where the customer terminal 100 is installed, but may be any place such as a place where the material to be purchased is processed or assembled, or a place of a warehouse.

  The desired material and the delivery destination information input by the consumer are transmitted to the material information server 400. The material information server 400 searches the procurement information database 420 and lists available material manufacturers for each desired material. (S500). Specifically, the material information server 400 refers to the sales material information, specifies a material maker that can provide the desired material for purchase, and generates a specified material maker list for each desired material for purchase. Thereafter, the procurement information calculation unit 430 estimates the cost and delivery date for each desired purchase material using the material maker list listed in the process of S500 and the procurement information stored in the procurement information database 420 ( S501).

  Here, it is assumed that the material is delivered using a small delivery service or a large delivery service provided by a distributor. In the small-lot delivery service, the cost of delivery can be calculated from, for example, the weight and size (size) of the package, the fare table for each area based on the address of the delivery source and the delivery destination, and the like.

  FIG. 20 shows an example of an area fare table for the small-lot delivery service. In the fare table for each area in FIG. 20, the area name of the delivery source and the area name of the delivery destination form a row and a column, and a column where the area name described in the row and the area name described in the column intersect. Stores the delivery cost. For example, when the delivery source is “Hokkaido” and the delivery destination is “Hokkaido”, a delivery cost of 594 yen is generated. Similarly, when the delivery source is “Hokkaido” and the delivery destination is “Okinawa”, a delivery cost of 1,188 yen is generated. The fare table by area is stored in the procurement information database 420.

  FIG. 21 shows an example of a fee table for each transport size. In the fare table of FIG. 21, the fares in the same area are defined for each of the maximum dimensions of the length, width, and depth of the cargo. For example, if "Hokkaido" is delivered as the same area, a delivery cost of 1,470 yen will be incurred if the maximum dimension is 60 cm or more and less than 80 cm. Similarly, if the maximum dimension is 80 cm or more and less than 100 cm, a delivery cost of 1,890 yen is generated. The charge table for each size is also stored in the procurement information database 420.

  The administrator of the material information server 400 may input the area-based fare table and the size-based fare table distributed, disclosed, and made public to the procurement information database 420 from a logistics company that provides a small-volume delivery service. Further, the material information server 400 or a computer cooperating with the material information server 400 obtains a fare table by area or a price table by size from a website or the like of each distributor by a computer program such as a search engine, and obtains a procurement information database 420. May be stored.

  The procurement information calculation unit 430 estimates the delivery cost by referring to the area-based fare table and the size-based fee table stored in the procurement information database 420 when delivering purchased materials using the small-volume delivery service. can do.

  On the other hand, when using the mass delivery service, it is possible to use a mixed flight, a charter flight, a home delivery service, or the like by distance. In the case of a mass delivery service, it may be difficult to calculate a charge by using a table. For example, an estimate service provided by a distribution company is used. The estimation service provided by the distributor can be used, for example, via an API (Application Programming Interface). FIG. 22 is an example of a data flow when the procurement information calculation unit 430 uses the estimation service via the API.

  The procurement information calculation unit 430 acquires delivery information such as the size and weight of the material desired to be purchased, the address of the delivery source, and the address of the delivery destination from the procurement information database 420 (Z1), and provides an estimation service via the network 200. Connect to the website of the distributor. Then, the material information server 400 searches for and uses the route / distance search service Z2 and the delivery company estimation service Z3 provided by the distribution company. When using the route / distance search service Z2 and the delivery company estimation service Z3, delivery information such as a delivery address of a material, a delivery address, a size and a weight is input.

  In the example of FIG. 22, the distance information on the route between the delivery source and the delivery destination obtained based on the route / distance search service Z2 is included in the mixed flight estimation master and the charter flight estimation master (Z4) for the fare estimation. Is entered. When the route / distance search service Z2 is not provided, the delivery cost based on the delivery information is estimated via the delivery company estimation service Z3. The procurement information calculation unit 430 acquires estimation information estimated through the route / distance search service Z2 and the delivery company estimation service Z3 provided by the distribution company. The estimation information includes a delivery cost and a delivery date (delivery period).

  The purchase cost of the desired material and the period until the shipment are specified from the sales material information stored in the procurement information database 420. Further, as described with reference to FIGS. 20 to 22, the delivery cost and delivery period of the purchased material are specified based on information (fare table, estimation service, etc.) provided by the distributor.

  Therefore, the procurement information calculation unit 430 calculates the procurement cost for each desired purchase material by adding the specified purchase cost and delivery cost. In the same manner, the delivery time is calculated for each desired purchase material by adding the period up to shipping and the delivery period for each desired purchase material. The procurement information calculation unit 430 generates a supplier selection support table for each desired purchase material based on the calculated procurement cost and delivery date (S501).

  The material information server 400 displays the material procurement plan table PT based on the supplier selection support table generated in S501 in order to display an estimate of the cost / delivery date of the material to be ordered on the customer terminal 100 (S153). Created and transmitted to the consumer terminal 100.

  FIG. 23 shows an example of a material procurement plan table PT for presenting a plurality of material procurement plans as options. In the material procurement plan table PT, as an example of a material procurement plan, a plan A that prioritizes performance (emphasis), a plan B that prioritizes cost (emphasis), and a plan C that prioritizes delivery time (emphasis) are presented in parallel. In this material procurement plan, the material to be procured is “stainless steel (SUS304)”.

  In preparing the procurement plan, the material information server 400 searches for the actual information stored in the procurement information database 420 for the material desired to be purchased (here, stainless steel (SUS304)). For example, the frequency (number of times) of purchase results is extracted from the searched result information. The material information server 400 compares the extracted performance frequencies and specifies the material and the material maker with the highest performance frequency (the number of times of the transaction performance). Here, the extracted performance frequency may be weighted according to the amount of the transaction performance. The material information server 400 presents, as a plan A (emphasis on performance), information of a record including the purchased material having a high track record in the supplier selection support table for each desired material for purchase.

  Similarly, the material information server 400 extracts the record with the lowest procurement cost from the supplier selection support table for each desired purchase material, and presents the information of the extracted record as Plan B (priority on the price). Further, the material information server 400 extracts the record with the shortest delivery date from the supplier selection support table for each desired purchase material, and presents the information of the extracted record as Plan C (emphasis on the delivery date).

  In the material procurement plan table PT, a “material” column stores a material maker, a material name, dimensions, and the like to be purchased in each plan. The “material cost” column stores the material purchase cost presented in each plan. The “delivery cost” column stores the delivery cost of the material presented in each plan. The “total” column stores the procurement costs of the materials presented in each plan. The “delivery date (business day)” column stores the delivery date of the material presented in each plan. In the “distribution” column, the magnitude of the distribution amount in the transaction is displayed. Any method may be used as long as the magnitude of the distribution amount can be displayed. In this embodiment, the information is displayed in three stages (large, medium, and small). The "Global Characteristics" column displays the country of origin of the material and the standard to which the material conforms. By indicating the country of origin of the material, it is possible to confirm in advance the procurement route when the customer produces or manufactures in a country other than Japan, or if the customer wants to obtain the material outside of Japan, It can also be purchased directly. Indication of the conforming standard allows one to determine whether a product using the material can be exported to another country. Standards include international standards such as ISO (International Organization for Standardization), JIS (Japanese Industrial Standards), AISI (American Iron and Steel Institute), DIN (Deutsche Industrie Normen), RoHS (Restriction on Hazardous Substances), and domestic standards. .

  In the form of the material procurement plan table PT illustrated in FIG. 23, the consumer who has been presented with a plurality of material procurement plans relatively compares the content presented in each plan and determines a desired material procurement plan. It is possible to make a selection and determine the procurement of materials. The material information server 400 may display a graph for comparing the procurement cost and delivery date of each plan in addition to the material procurement plan PT. The customer can visually check and compare the procurement costs and the delivery dates of the respective plans arranged in parallel in the material procurement plan table PT in a visualized form.

  The consumer selects one or more of the plurality of plans presented on the consumer terminal 100, and requests the material information server 400 to create an estimate (quotation) for the selected plan (S154). For example, the consumer can relatively compare the presented cost and delivery date information for each procurement plan, specify a cost condition close to the purpose, and specify a procurement plan that matches the delivery date condition, and make a request to create an estimate. . If you select multiple plans, you will get a phase estimate.

  When the customer requests the material information server 400 for an estimate (in the case of phase estimation), the material information server 400 acquires the materials of the procurement plan selected for the terminals 2000a, 2000b, and 2000c of the material manufacturers M1, M2, and M3. Send the information necessary for estimation, such as quantity, dimensions, delivery date, etc., and request an estimate. The material manufacturers M1, M2, and M3 receive the request for quotation and provide necessary information. Further, the material information server 400 receives the answers of the material manufacturers M1, M2, and M3, and inquires of the distribution companies L1 and L2 about the delivery cost to the delivery destination and the delivery date and time delivered to the delivery destination. The distributors L1 and L2 reply the delivery cost and the delivery date and time to the material information server 400. The procurement information calculation unit 430 of the material information server 400 creates a quote based on the responses from the material manufacturers M1, M2, and M3 and the logistics companies L1 and L2. As a result, a highly accurate quote can be created.

  FIG. 24 is an example of a created estimate. FIG. 24 is an example of an estimate corresponding to Plan A in the material procurement plan table PT. Similarly, quotations corresponding to other plans can be created. In creating an estimate, for example, it is assumed that a template of the estimate as shown in FIG. 24 is stored in advance in the procurement information database 420 of the material information server 400. The material information server 400 refers to the template of the estimate and describes items necessary for creating the estimate.

  In the quote shown in FIG. 24, the name of the customer (company name, person in charge) requesting the creation of the quote is described as the destination of the quote. Further, the material name and the delivery date of Plan A shown in the material procurement plan table PT are described as the “subject” and “delivery date” of the quote, respectively. Then, the contents corresponding to Plan A are described as “Description”, “Quantity”, “Unit Price”, and “Amount”. In “Notes”, the delivery source, delivery destination address, and the like of the materials used for calculating the delivery cost of Plan A are described.

  The material information server 400 outputs the created estimate to the consumer terminal 100. The estimate is output in, for example, an HTML (Hyper Text Markup Language) format or a PDF (Portable Document Format) format. The quote illustrated in FIG. 24 is presented to the consumer via a display screen such as an LCD provided in the consumer terminal 100. The customer places an order after seeing the quote (S155). When the customer places an order, the material information server 400 places an order including necessary information such as material, quantity, and delivery date to the material maker and the distributor of the material ordered by the consumer via the material maker terminal 2000 and the distributor terminal 2500. Is issued (S503). As a result, the material information server 400 can provide a one-stop service that supports the determination of the selection and arrangement of a plurality of related suppliers for material procurement.

  Although the material specification information server 300 and the material information server 400 are separately described here for the purpose of clarifying the configuration, the material specification information server 300 also serves as the material information server 400, or the material information server 400 The information server 300 may also be used.

  Note that the present invention is not limited to the above-described embodiment, and includes various modifications. For example, the above-described embodiments have been described in detail for easy understanding of the present invention, and are not necessarily limited to those having all the configurations described above. Further, the processing methods described in the embodiments are also exemplarily illustrated for easily explaining the present invention, and are not limited to the above-described processing methods. Further, a part of the configuration of one embodiment can be replaced with the configuration of another embodiment, and the configuration of one embodiment can be added to the configuration of another embodiment. Also, for a part of the configuration of each embodiment, it is possible to add, delete, or replace another configuration.

100: customer terminal, 110: material use condition input unit, 120: specification item determination unit, 130: material search unit, 140: material selection unit, 200: communication line, 300: material specification information server, 310: condition decomposition unit , 320: specification information database system, 330: candidate output unit, 400: material information server, 410: material information database, 420: procurement information database, 430: procurement information calculation unit.

Claims (22)

A material specification information server connected to the terminal via a communication line and supporting selection of a material used for the product,
A registered word file for registering a plurality of registered words, a first mode correlation matrix file for registering correlation values between the plurality of registered words and a plurality of first modes relating to the product, a plurality of properties relating to the material, and the plurality of first modes; A property for registering a correlation value with the first mode-a data file for storing a first mode correlation matrix file;
A condition decomposition unit that generates a search term string from the material use conditions input from the terminal,
A specification information database that outputs a material property sequence from the search term sequence,
A candidate output unit that outputs a predetermined specification item candidate sequence from the material property sequence to the terminal,
The search word string has a plurality of pairs of registered words and their weights included in the registered word file,
The specification information database calculates the plurality of first mode evaluation values for the search word string using the search word string and the first mode correlation matrix file, and calculates the plurality of first mode evaluation values. Calculating a first evaluation value of the plurality of characteristics with respect to the search word string using the characteristic-first mode correlation matrix file; and the material including a pair of the plurality of characteristics and the first evaluation value of each of the plurality of characteristics. Generate a characteristic column,
The material specification information server, wherein the candidate output unit selects a predetermined number of characteristics having a high first evaluation value from the material characteristic sequence as specification item candidates of the specification item candidate sequence. In claim 1,
The data file stores a corpus file,
The condition decomposing unit decomposes the input material use condition into a word string including a plurality of words. If the word is a registered word included in the registered word file, the similarity is set to 1; Is not included in the registered word file, a plurality of registered words similar to the word are extracted using the corpus file, and the similarity between the extracted registered word and the word is calculated,
A material specification information server that normalizes the similarity of each registered word corresponding to the word string and uses the normalized similarity as the weight. In claim 2,
The material use condition is defined by a plurality of use conditions and a logical operator that combines each of the plurality of use conditions,
A material specification information server including, as the plurality of use conditions, the product in which the material is used, an environment in which the product is used, and a processing method applied to the material to manufacture the product. In claim 3,
The condition decomposition unit generates a plurality of search term strings from the plurality of use conditions included in the material use condition,
The specification information database calculates, for the plurality of search word strings, evaluation values of the plurality of first modes using the first mode correlation matrix file, and calculates the plurality of first mode evaluation values for each of the plurality of search word strings. By combining the evaluation values of the first mode in accordance with the logical operator, a plurality of first mode combined evaluation values for the plurality of search word strings are calculated, and the plurality of first mode combined evaluation values are calculated. And calculating a first evaluation value of the plurality of characteristics with respect to the plurality of search word strings using the characteristic-first mode correlation matrix file, and forming a pair of the plurality of characteristics and each of the first evaluation values. A material specification information server that generates the material characteristic sequence including the information. In claim 1,
The specification information database generates a first mode sequence in which the plurality of first modes and the evaluation values of the plurality of first modes for the search word sequence are paired,
The first mode column is a material specification information server output to the terminal. In claim 1,
The first mode is a material specification information server that categorizes development of the product using the material, manufacture of the product, and damage that occurs during use of the product. In claim 1,
The data file includes a second mode correlation matrix file for registering a correlation value between the plurality of registered words and a plurality of second modes related to the product, and a correlation between the plurality of characteristics related to the material and the plurality of second modes. A property for registering a value-storing the second mode correlation matrix file,
The specification information database calculates the evaluation values of the plurality of second modes for the search word string using the search word string and the second mode correlation matrix file, and calculates the evaluation values of the plurality of second modes. Calculating a second evaluation value of the plurality of characteristics with respect to the search word string using the characteristic-second mode correlation matrix file, and the material including a pair of the plurality of characteristics and each of the second evaluation values; Generate a characteristic column,
The material specification information server, wherein the candidate output unit selects a predetermined number of characteristics having a high second evaluation value from the material characteristic sequence as specification item candidates of the specification item candidate sequence. In claim 7,
The second mode is a material specification information server that categorizes the quality required for the product. In claim 7,
The first mode correlation matrix file and the second mode correlation matrix file are configured as one correlation matrix file, and the characteristic-first mode correlation matrix file and the characteristic-second mode correlation matrix file are one characteristic-mode. Material specification information server configured as a correlation matrix file. In claim 1,
The data file stores a reference material correlation matrix file for registering a correlation value between the plurality of registered words and a plurality of reference materials,
The specification information database calculates an evaluation value of the reference material for the search word string using the search word string and the reference material correlation matrix file, and the plurality of reference materials and the reference material for the search word string. Generate a reference material candidate sequence paired with the evaluation value of
The reference material candidate sequence is a material specification information server output to the terminal. In claim 10,
A material specification information server, wherein the first mode correlation matrix file and the reference material correlation matrix file are configured as one correlation matrix file. A material specification information server according to any one of claims 1 to 11,
A material selection support system comprising: a material information server having a material information database for storing characteristic values of a plurality of materials. In claim 12,
The material selection support system, wherein the material information server includes a procurement information database that stores procurement information necessary for procurement of the material. A material selection support method for supporting selection of a material used for a product,
Input the material usage conditions from the terminal to the material specification information server,
The material specification information server registers in advance a plurality of registered words, a correlation between the plurality of registered words and a plurality of first modes relating to the product, and a correlation between a plurality of characteristics relating to the material and the plurality of first modes. ,
Generate a search word string including a plurality of pairs of the registered word and its weight from the material use conditions,
Based on the weight of the search word string and the correlation between the plurality of registered words and the plurality of first modes related to the product, calculate the evaluation values of the plurality of first modes for the search word string,
Based on the evaluation values of the plurality of first modes for the search word string and the correlation between the plurality of properties and the plurality of first modes, calculate a first evaluation value of the plurality of properties for the search word string,
A material selection support method for displaying a predetermined number of characteristics having a high first evaluation value as specification item candidates on the terminal. In claim 14,
As the material use conditions, a plurality of use conditions and a logical operator that combines each of the plurality of use conditions are input,
A material selection support method including, as the plurality of use conditions, the product in which the material is used, an environment in which the product is used, and a processing method applied to the material to manufacture the product. In claim 14,
A material selection support method for displaying, on the terminal, a first mode having a higher evaluation value of the first mode among the plurality of first modes. In claim 14,
The material specification information server registers in advance the correlation between the plurality of registered words and the plurality of reference materials,
Based on the weight of the search word string and the correlation between the plurality of registered words and the plurality of reference materials, calculate the evaluation value of the plurality of reference materials for the search word string,
A material selection support method for displaying, on the terminal, a reference material having a high evaluation value of the reference material among the plurality of reference materials. In claim 14,
Inputting specification values for at least one or more specification items determined from the specification item candidates displayed on the terminal,
A material selection support method for searching for a material by inputting a material search condition including the specification item and the specification value to a material information server. In claim 17,
At least one or more specification items determined from the specification item candidates displayed on the terminal and a reference material having a high evaluation value of the reference material are input to a material information server,
A material selection support method, wherein the material information server outputs a specification value of a reference material having a high evaluation value of a reference material corresponding to the specification item to the terminal as a reference specification value. In claim 14,
The first mode is a material selection support method that categorizes damages that occur during development of the product using the material, manufacture of the product, and use of the product. In claim 18,
The material information server has a procurement information database that stores procurement information necessary for procurement of the material,
A material selection support method for creating at least one or more procurement plans based on the procurement information database for purchase desired materials input from the terminal. In claim 21,
The material information server receives a request from the terminal and creates a quote based on the procurement plan.

Images