KR102306917B1 - Method and apparatus for determining the optimal chemical process - Google Patents

Abstract

The present invention relates to a method and apparatus for determining an optimal chemical process. More specifically, the method for determining an optimal chemical process includes the steps of: constructing a database including a plurality of process recipes extracted from thesis data; selecting a first criterion from among a plurality of criteria for determining an optimal process from among the plurality of process recipes; and determining the optimal process by using the first criterion.

Description

Method and apparatus for determining the optimal chemical process

The present invention relates to a method and apparatus for determining the optimal chemical process of system performance data. More particularly, it relates to a method and apparatus for determining an optimal chemical process capable of determining an optimal process recipe from among a plurality of process recipes extracted from thesis data.

It is known that there are more than 100 million chemical substances that have not been revealed through prior research, and it is known that there are about 1,060 possible combinations of compounds to make new chemical substances (Kishimoto, Buesser, & Botea, 2018). Discovering new chemicals is a high-value-added task that secures basic resources that can be used in various products, and is a field that is being studied with great interest by various industries.

Conventionally, a method of manually searching for possible chemical substances based on expert knowledge has been used, but recently, research using natural language processing technology is being conducted.

In particular, there was an attempt to extract meaningful scientific knowledge from unstructured data and apply it to actual R&D. Attempts to utilize scientific and technological text information for actual research are actively underway, mainly in the fields of materials, processes, and chemistry.

However, unlike other technologies, when extracting information from thesis data including chemical formulas or thesis data describing processes in the chemical field, accurate information could not be extracted due to limitations of existing natural language processing technologies.

Also, in the prior art, there was no specific means for effectively utilizing the automatically extracted chemical process or chemical field information, except for extracting and storing information from the thesis data.

Therefore, there is a need for a method to effectively utilize the information extracted from the thesis data.

Laid-open Patent Publication KR 10-2010-0080904 (published on July 13, 2010)

The technical problem to be solved by the present invention is to provide an optimal chemical process determination method and apparatus capable of automatically extracting procedural knowledge from scientific literature and automatically extracting chemical processes using natural language processing technology.

Another technical problem to be solved by the present invention is to provide a method and apparatus for determining an optimal chemical process capable of extracting an optimal process from among chemical processes automatically extracted from thesis data.

In addition, another technical problem to be solved by the present invention is to provide a method and apparatus for determining an optimal chemical process capable of determining an optimal chemical process from among a plurality of process recipes according to a criterion for determining an optimal process.

In addition, another technical problem to be solved by the present invention is to provide an optimal chemical process determination method and apparatus capable of automatically determining and extracting a chemical process optimized for a necessary purpose by various criteria and a method for determining an optimal process according thereto will do

The technical problems of the present invention are not limited to the technical problems mentioned above, and other technical problems not mentioned will be clearly understood by those skilled in the art from the following description.

The method for determining an optimal chemical process according to an embodiment of the present invention for solving the above problems includes: building a database including a plurality of process recipes extracted from thesis data; determining an optimal process from among the plurality of process recipes selecting a first criterion from among a plurality of criteria for

In an embodiment, the step of selecting a first criterion from among a plurality of criteria for determining an optimal process from among the plurality of process recipes includes determining the amount of intermediates generated in the process of performing each of the process recipes from among the plurality of criteria. The method may include selecting a first criterion determined based on the first criterion.

In an embodiment, the determining of the optimal process by using the first criterion includes: identifying an intermediate produced in the process of performing each of the plurality of process recipes; and the intermediate material from among the plurality of process recipes. It may include determining a process recipe having the smallest amount as the optimal process.

In an embodiment, the selecting of a first criterion among a plurality of criteria for determining an optimal process from among the plurality of process recipes includes selecting a first criterion according to the implementation difficulty of an environment in which the plurality of process recipes are performed. may include steps.

In an embodiment, the determining of the optimal process using the first criterion includes determining a temperature for implementing a process to be performed with respect to a first process recipe among the plurality of process recipes, and the temperature is set in advance. It may include determining the first process recipe as the optimal process when it corresponds to the set temperature.

In one embodiment, the selecting of a first criterion among a plurality of criteria for determining an optimal process from among the plurality of process recipes includes selecting a first criterion according to reliability of a process performed of the plurality of process recipes. may include

In an embodiment, the determining of the optimal process using the first criterion includes repeating the process of performing the plurality of process recipes to determine success rates of the plurality of process recipes, and the plurality of process recipes. and determining the process recipe having the highest success rate among the process recipes as the process recipe having the high reliability of the first criterion, and determining the process recipe as the optimal process.

In an embodiment, the step of selecting a first criterion from among a plurality of criteria for determining an optimal process from among the plurality of process recipes includes a first criterion according to the cost of a material used in the process of performing the plurality of process recipes. It may include the step of selecting.

In an embodiment, the determining of the optimal process by using the first criterion includes: identifying a material used in a process for performing each of the plurality of process recipes; and a process for performing each of the plurality of process recipes and determining, as the optimal process, a process recipe having the lowest total sum of costs of the raw materials among the plurality of process recipes based on the total sum of costs of the raw materials.

In an embodiment, the determining of the optimal process using the first criterion comprises: scoring each of the plurality of process recipes based on a weight assigned to the first criterion; It may include the step of determining an optimal process according to the.

In an embodiment, the scoring each of the plurality of process recipes based on the weight assigned to the first criterion may include a weight assigned to the first criterion and a weight assigned to the first criterion and different from the first criterion among the plurality of criteria. The method may include scoring each of the plurality of process recipes using a sum of weights assigned to the second criterion.

In an embodiment, scoring each of the plurality of process recipes using a score obtained by adding a weight assigned to the first criterion and a weight assigned to the second criterion may include: and excluding the process recipe from the optimal process when there is a process recipe having a score of 0 according to a score or a weight assigned to the second criterion.

In an embodiment, the determining of the optimal process by using the first criterion includes determining the optimal process by using a first process recipe selected based on the first criterion among the plurality of process recipes. may include

In an embodiment, the step of constructing a database including a plurality of process recipes extracted from the thesis data includes extracting a process sentence from the thesis data, and identifying a process verb for generating a target material in the first process sentence and extracting a subject and an object corresponding to the process verb, when the process verb is changed to an adjective in the second process sentence and the subject and the object are not identified, the target material generated in the first process sentence is used as a subject It may include the step of determining.

In one embodiment, the step of constructing a database including a plurality of process recipes extracted from the thesis data includes learning to analyze the meaning of the text included in the thesis data to obtain a synonym list database for text having the same meaning. constructing , and extracting a process sentence from the thesis data using the synonym list database.

An apparatus for determining an optimal chemical process according to another embodiment includes a processor, a network interface, a memory executed by the processor to load a computer program, and a storage for storing the computer program, wherein the computer program comprises: an instruction for constructing a database including a plurality of process recipes extracted from data, an instruction for selecting a first criterion from among a plurality of criteria for determining an optimal process from among the plurality of process recipes, and the first criterion It may include instructions for determining the optimal process using

A computer-readable recording medium according to another embodiment is a computer program for determining an optimal execution process, comprising computer program instructions executable by a processor, when the computer program instructions are executed by a processor of a computing device, Building a database including a plurality of process recipes extracted from thesis data, selecting a first criterion from among a plurality of criteria for determining an optimal process from among the plurality of process recipes, and using the first criterion A computer program may be recorded for performing operations, including determining the optimal process.

1 is an exemplary diagram illustrating an operation of an apparatus for determining an optimal chemical process according to an embodiment of the present invention.
2 is a flowchart of a method for determining an optimal chemical process according to another embodiment of the present invention.
3 is an exemplary diagram illustrating a process of constructing a process recipe database from thesis data.
4 is a diagram illustrating an example in which a subject recognition error occurs in the process of extracting a process recipe from thesis data.
5 is a diagram illustrating an example in which an object recognition error occurs in a process of extracting a process recipe from thesis data.
6 is a flowchart for specifically explaining steps S200 and S300 of FIG. 2 .
7 is a flowchart illustrating various embodiments of step S300 of FIG. 2 .
8 is a hardware configuration diagram of an apparatus for determining an optimal chemical process according to an embodiment of the present invention.

Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings. Advantages and features of the present invention and methods of achieving them will become apparent with reference to the embodiments described below in detail in conjunction with the accompanying drawings. However, the technical spirit of the present invention is not limited to the following embodiments, but may be implemented in various different forms, and only the following embodiments complete the technical spirit of the present invention, and in the technical field to which the present invention belongs It is provided to fully inform those of ordinary skill in the art of the scope of the present invention, and the technical spirit of the present invention is only defined by the scope of the claims.

In adding reference numerals to the components of each drawing, it should be noted that the same components are given the same reference numerals as much as possible even though they are indicated on different drawings. In addition, in describing the present invention, if it is determined that a detailed description of a related known configuration or function may obscure the gist of the present invention, the detailed description thereof will be omitted.

Unless otherwise defined, all terms (including technical and scientific terms) used herein may be used with the meaning commonly understood by those of ordinary skill in the art to which the present invention belongs. In addition, terms defined in a commonly used dictionary are not to be interpreted ideally or excessively unless clearly defined in particular. The terminology used herein is for the purpose of describing the embodiments and is not intended to limit the present invention. As used herein, the singular also includes the plural unless specifically stated otherwise in the phrase.

In addition, in describing the components of the present invention, terms such as first, second, A, B, (a), (b), etc. may be used. These terms are only for distinguishing the components from other components, and the essence, order, or order of the components are not limited by the terms. When a component is described as being “connected”, “coupled” or “connected” to another component, the component may be directly connected or connected to the other component, but another component is between each component. It should be understood that elements may be “connected,” “coupled,” or “connected.”

As used herein, “comprises” and/or “comprising” refers to the presence of one or more other components, steps, operations and/or elements mentioned. or addition is not excluded.

Hereinafter, some embodiments of the present invention will be described in detail with reference to the accompanying drawings.

1 is an exemplary diagram illustrating an operation of an apparatus 100 for determining an optimum chemical process according to an embodiment of the present invention.

Referring to FIG. 1 , the apparatus 100 for determining an optimal chemical process according to the present embodiment may extract a process recipe from a plurality of papers to determine an optimal process recipe.

Specifically, the apparatus 100 for determining the optimal chemical process may pre-process a plurality of thesis data and extract a process recipe included in the plurality of thesis data by machine learning the pre-processed thesis data.

In the present specification, the process recipe includes a chemical formula, a chemical reaction formula, a plurality of chemical reaction formulas, a chemical formula to generate a target material, organic oxides synthesis data, the title and metadata of a material science journal article, and a Zeolite synthesis dataset , may include sentences, formulas, symbols, etc. for other chemical reactions, but is not limited thereto.

The optimal chemical process determination apparatus 100 pre-processes the unstructured text from the thesis data into structured data, and extracts entities for substances or materials through natural language processing from the pre-processed structured data using a model learned through machine learning (NER) and relationship extraction. The apparatus 100 for determining the optimal chemical process may extract a plurality of process recipes through the above operation.

The apparatus 100 for determining the optimal chemical process may input a criterion for determining the optimal process. The criterion for determining the optimal process is a target according to the process recipe, and may be an externally input or selected criterion. For example, in the apparatus 100 for determining the optimal chemical process, a criterion for determining an optimal process of a process procedure criterion, an environmental criterion, a reliability criterion, or a cost criterion may be input or selected.

The optimal chemical process determination apparatus 100 may determine an optimal process recipe based on the above criteria. The optimal chemical process determination apparatus 100 may determine different optimal process recipes according to characteristics of each criterion.

In this case, the apparatus 100 for determining the optimal chemical process may determine the optimal process by using the first process recipe selected based on the first criterion among the plurality of process recipes. Here, the first process recipe may mean any one process of generating a target material from among a plurality of process recipes, or at least a plurality of processes, but is not limited thereto.

The apparatus 100 for determining an optimal chemical process according to an embodiment of the present invention may determine an optimal chemical process from among a plurality of process recipes automatically extracted from thesis data as long as a criterion for determining the optimal process is input.

So far, the operation of the apparatus 100 for determining the optimum chemical process according to an embodiment of the present invention has been schematically described with reference to FIG. 1 . Hereinafter, an operation of the method for determining an optimal chemical process according to an embodiment of the present invention will be described with reference to FIGS. 2 to 8 . This embodiment may be performed by a computing device. For example, the computing device may be the apparatus 100 for determining the optimal chemical process described with reference to FIG. 1 . In describing the present embodiment, descriptions of subjects performing some operations may be omitted. In this case, the performing subject is the computing device.

2 is a flowchart of a method for determining an optimal chemical process according to another embodiment of the present invention, and FIG. 3 is an exemplary diagram illustrating a process of constructing a process recipe database from thesis data.

Referring to FIG. 2 , a database including a plurality of process recipes extracted from thesis data in step S100 may be built. The database including a plurality of process recipes may be a database in which chemical formulas for substances are stored.

As shown in FIG. 3 , a database including a plurality of process recipes may store a chemical process extracted from chemical thesis data through a natural language processing-based model. In this step, the thesis data may be chemical thesis data in which the chemical formula is disclosed, but is not limited thereto.

A database including a plurality of process recipes may automatically generate a recipe or store an optimized recipe through the recipe automatic creation framework and the recipe optimization framework. In addition, a database including a plurality of process recipes may be linked to a compound DB, and may be managed by an expert-based verification support system of a chemical process.

Referring back to FIG. 2 , when the database is built in step S100, a pre-processing step may be performed first. Specifically, in the content acquisition stage, which is the first stage, journal articles whose authors are judged to be related to the field of materials science may be collected based on chemistry-related thesis data. The original manuscript collection is carried out in the form of collecting HTML/XML files of papers published after 2000 using the scrapy tool, and the collected information can be stored in MongoDB.

In the Paragraphs classification stage as the second stage, a two-stage approach can be used to find paragraphs related to solid-state synthesis in the thesis data.

In the topic labeling step through clustering, clustering may be performed based on the keywords of the experiment and related paragraphs, and topic labeling may be performed based on keywords constituting each cluster.

In the step of classifying the paragraph topic through the random forest classifier, the main topic of the paragraph can be predicted with the classifier after learning the random forest classifier to match the main topic among the various topics constituting the paragraphs. Solid-state synthesis), hydrothermal synthesis, sol-gel precursor synthesis, and n/a were identified as none, and it can be configured to study 1000 paragraphs for each label.

In step 3, the synthesis recipe extraction step, among the paragraphs classified above, the paragraphs related to solid-phase synthesis were used for (1) material entity recognition, (2) synthesis operations. , (3) mixing and heating conditions, and (4) balancing equations can be extracted.

More specifically, in the material entity name recognition stage, BiLSTM-CRF, which combines Bi-directional LSTM (Long-short term memory) neural network and conditional random field layer, and Word2Vec model to TARGET, PRECURSOR the entity names of words in chemical papers , a recognizer that recognizes as one of OTHER can be trained, and 834 solid-phase synthetic paragraphs extracted from 750 papers can be utilized for learning.

The synthesis method utilizes a neural network, Word2Vec, and SpaCy, and in terms of the synthesis method, a classifier that classifies words in a sentence into one of six labels such as NOT OPERATION, MIXING, HEATING, DRYING, SHAPING, and QUENCHING can be learned.

In the combination and heating condition setting step, using regular expressions to find words related to combination and heating conditions, such as time and temperature, labeling can be carried out so that it can be utilized as detailed information of the synthesis method labeled HEATING and MIXING.

Based on the information extracted in the chemical reaction equation balancing step, the process process can be expressed and stored in the form of a chemical formula.

In the fourth step, the dataset generation step, data can be extracted and stored from a large amount of papers by using the previously learned chemical process information extractor. From 4,204,170 papers, 6,218,136 paragraphs related to the experimental section can be secured, of which 188,198 paragraphs related to inorganic substances can be secured. Of these, 53,538 paragraphs related to solid phase synthesis are again classified, and they are used to secure a dataset related to chemical synthesis process recipes. was found to be

Thereafter, using the pre-processed data, paragraph data information for the synthesis of solid-state, hydrothermal, and sol-gel precursors, the name of the inorganic compound (Target), the material of the inorganic compound (Precursor), the synthesis operation (Operation) and the synthesis operating conditions An extraction model can be trained using a recipe dataset consisting of (Condition).

In addition, when the data set is built, in some papers, the chemical formula is not unified, and the chemical formula is expressed in various forms such as empirical formula, molecular formula, and abbreviated structural formula. A problem that cannot be annotated may occur.

Therefore, in the case of paragraph data in which the chemical formula of the thesis and the reference dataset are not unified, only 2,852 paragraph data out of a total of 3,105 paragraph data were used as training data.

11,345 sentences were constructed by separating paragraph data into sentences using the 'NLTK (Natural Language Toolkit)' Python external library1) for learning the entity name recognition model. The library shows high performance for unstructured texts in general domains, but in domains containing some specialized contents, there were cases in which sentences were incorrectly separated by understanding the context incorrectly. Therefore, there were cases where sentence separation did not proceed normally due to the nature of the inorganic compound formula.

The size of the finally constructed entity name recognition model training data is 2,852 cases of training paragraph data and 11,345 sentences of learning sentence data. It was composed of sentences and 2,016 sentences of test data.

A model was trained using this training data, and in this embodiment, a bidirectional LSTM-CRF combining Bi-LSTM and CRF was used as a model, and as an input, it can be embedded into BioBERT, ClinicalBERT, and SciBert models based on the BERT model. have.

The extraction model used in this step can output the compound name of the inorganic compound (Target), the compound material (Precursor), and the synthesis operation (Operation) when the thesis data is input.

As a result of the verification of the extraction model, the F1-score of BioBERT, a biomedical language expression model, was measured to have the highest performance as 80.78, and it was found that biomedical language was most closely related to chemistry-related fields.

An example of resolving an error occurring in the process of extracting a process recipe using the extraction model will be described with reference to FIGS. 4 and 5 . 4 is a diagram illustrating an example in which a subject recognition error occurs in the process of extracting a process recipe from thesis data, and FIG. 5 is a diagram illustrating an example in which an object recognition error occurs in the process of extracting a process recipe from thesis data.

As shown in Figure 4, although the word of calcined was used as a verb in the previous sentence, it is rarely used as an adjective in a later sentence, so there is a problem that occurs because the model cannot learn correctly. In the first sentence, dried and calcined are 'dried and Although ‘calcined’ indicates a verb-form operation (Operation), The calcined products in the second sentence means ‘calcined object’ as a noun, so there is a problem that it is used in a different form from the operation. .

In an embodiment for solving this problem, when a process sentence is extracted from the thesis data, a process verb generating a target material may be identified from the first process sentence, and a subject and an object corresponding to the process verb may be extracted.

When the process verb is changed to an adjective in the second process sentence and the subject and the object are not identified, the target material generated in the first process sentence may be determined as the subject.

For example, the extraction model may not extract a sentence having a structure composed of a general subject-verb-object word, but may recognize a verb from a paragraph determined to be fair. In this case, if the verb is a target fair verb (operation) and a subject (target) and an object (precursor) exist in the corresponding sentence, the object corresponding to the subject and precursor corresponding to the process verb and the target may be extracted.

Thereafter, when the adjective "milled", not a verb, is recognized and the subject and object are not recognized, the extraction model may determine that there is no object and treat the output of the preceding sentence as the subject. However, when the extraction model identifies that a new object appears with a verb such as "add" or "combine", it can treat the output of the previous sentence as a subject and associate it with a new object.

As shown in FIG. 4 , it is a problem that occurs when the name of the inorganic compound (Target) is expressed as an abbreviation, and in the case of BNT, it is difficult to predict because it is different from the general chemical formula.

In another embodiment for solving this problem, learning to analyze the meaning of the text included in the thesis data may be performed to build a synonym list database for the text having the same meaning. Thereafter, a process sentence may be extracted from the thesis data using a synonym list database.

In this embodiment, the process recipe can be extracted using the synonym list database only in the case of synonyms existing in other paragraphs.

In object recognition, unlike general relationship extraction, if even one sentence is omitted, an error occurs in the whole. This extraction model can extract process recipes with higher performance because it uses a synonym list database for synonyms that exist in other paragraphs.

Referring back to FIG. 2 , a first criterion among a plurality of criteria for determining an optimal process among a plurality of process recipes may be selected in step S200, and the optimal process may be determined using the first criterion in step S300 .

In order to describe steps S200 and S300 in detail, reference is made to FIG. 6 . 6 is a flowchart for specifically explaining steps S200 and S300 of FIG. 2 .

When step S200 is performed, steps S210 to S240 may be selectively performed. In step S210, a first criterion determined based on the amount of intermediates generated in the process of performing each of the process recipes may be selected from among the plurality of criteria, and in step S220, the implementation difficulty of the environment in which the plurality of process recipes is performed A first criterion according to the may be selected. In step S230, the first criterion according to the reliability of the process of performing the plurality of process recipes may be selected, and in step S240, the first criterion according to the cost of the material used in the process of performing the plurality of process recipes may be selected. Step S300 may be performed through the criteria in step S200.

Specifically, when the first criterion determined based on the amount of intermediates generated in each execution process is selected in step S210, intermediates generated in the process of performing each of a plurality of process recipes are identified in step S310, and a plurality of Among the process recipes of , a process recipe having the smallest amount of the intermediate material may be determined as an optimal process.

For example, when process recipes A, B, and C exist, 1 mg of the intermediate is produced in the process of performing process recipe A, and 5 mg of the intermediate is produced in the process of performing process recipe B, and process recipe C In the case where 10 mg of an intermediate is produced in the course of performing

When the first criterion according to the implementation difficulty of the environment in which the plurality of process recipes is performed is selected in step S220, a temperature for implementing each of the plurality of process recipes is determined in step S320, and among the plurality of process recipes A process recipe corresponding to a preset temperature for implementing a process of performing a plurality of process recipes may be determined as an optimal process. Here, the preset temperature may be a preset temperature such as 0 degrees, 10 degrees, or 20 degrees, a temperature set in a range such as 10 to 20 degrees, or a temperature set to room temperature, room temperature, or the like. In this case, the process recipe closest to the preset temperature may be determined as the optimal process.

For example, an example of a case where process recipes A, B, and C exist and the preset temperature is 20 to 25 degrees will be described. At this time, in the case of 20-25 degrees, the lowest level of difficulty is set to level 1, in the case of 15-19 degrees and 26-29 degrees, the difficulty level is set to higher than level 1, and in the case of 10-14 degrees and 30-34 degrees, it is set to level 2 It can be set to level 2 with high difficulty.

If the temperature at which the process recipe A is performed is 23 degrees, the temperature at which the process recipe B is performed is 17 degrees, and the temperature at which the process recipe C is performed is 33 degrees, the implementation difficulty of the process recipe A is level 1, the The implementation difficulty may be determined as level 2, and the implementation difficulty level of the recipe C may be determined as level 3. Accordingly, the process recipe A having the lowest implementation difficulty of level 1 may be determined as the optimal process. When the first criterion according to the reliability of the execution process of the plurality of process recipes is selected in step S230, among the plurality of process recipes in step S330 The process recipe with the highest success rate may be determined as the process recipe having the high reliability of the first criterion, and thus the process recipe may be determined as the optimal process.

When the first criterion according to the cost of the material used in the process of performing the plurality of process recipes is selected in step S240, the cost of the material is based on the sum of the costs of the material in the process performed for each of the plurality of process recipes in step S340. The process recipe with the lowest sum may be determined as the optimal process. In an embodiment, the total sum of the cost of the material may mean the purchase cost of the material, the management cost, the experiment execution cost, the material procurement cost, and the like.

For example, when process recipes A, B, and C exist, process recipe A is 1 million won, process recipe B is 5 million won, and process recipe C is input when the total cost of materials used in the process to be performed is input If is 10 million won, A with the smallest total cost of materials may be determined as the optimal process.

The optimal chemical process determination method according to another embodiment of the present invention has the advantage of automatically determining and extracting a chemical process optimized for a necessary purpose by the various criteria and the method for determining the optimal process according to the above-mentioned various criteria.

7 is a flowchart illustrating various embodiments of step S300 of FIG. 2 .

When step S300 is performed, steps S350 and S360 may be performed to determine an optimal chemical process.

Specifically, when step S350 is performed, scoring may be performed for each of the plurality of process recipes based on the weight assigned to the first criterion.

That is, a weight may be assigned to the first criterion, and scoring of the process recipe may be performed by reflecting the corresponding weight.

Additionally, multiple process recipes may be scored based on multiple criteria. In this case, the plurality of criteria may be at least one step or a combination of a plurality of steps performed in steps S210 to S240. In addition, each of the plurality of process recipes may be scored using a score obtained by adding the weight assigned to the first criterion and the weight assigned to the second criterion.

For example, if the weight assigned to the first criterion is 0.7 and the weight assigned to the second criterion is 0.3, the corresponding process recipe may be scored by reflecting the weight of the first criterion and the weight of the second criterion together. have.

In this case, when a process recipe having a score of 0 according to a weight assigned to the first criterion or a score according to a weight assigned to the second criterion exists, the process recipe having a score of 0 may be excluded from the optimal process. Accordingly, in the present embodiment, it is necessary to quickly exclude or exclude a process recipe having a score of 0 to increase the operation speed. This operation can prevent unnecessary operation and delay of the entire operation.

So far, the optimal chemical process determination method described herein has been described with reference to FIGS. 1 to 8 .

8 is a hardware configuration diagram of an apparatus for determining an optimal chemical process according to an embodiment of the present invention.

As shown in FIG. 8 , the computing device 500 loads one or more processors 510 , a bus 550 , a communication interface 570 , and a computer program 591 executed by the processor 510 . It may include a memory 530 and a storage 590 for storing the computer program (591). However, only the components related to the present invention are illustrated in FIG. 8 . Accordingly, a person skilled in the art to which the present invention pertains can see that other general-purpose components other than the components shown in FIG. 8 may be further included.

The processor 510 controls the overall operation of each component of the computing device 500 . The processor 510 includes at least one of a central processing unit (CPU), a micro processor unit (MPU), a micro controller unit (MCU), a graphic processing unit (GPU), or any type of processor well known in the art. may be included. In addition, the processor 510 may perform an operation on at least one application or program for executing the method/operation according to various embodiments described herein. Computing device 500 may include one or more processors.

The memory 530 stores various data, commands, and/or information. Memory 530 may load one or more programs 591 from storage 590 to execute methods/operations according to various embodiments described herein. An example of the memory 530 may be a RAM, but is not limited thereto.

The bus 550 provides communication between components of the computing device 500 . The bus 550 may be implemented as various types of buses, such as an address bus, a data bus, and a control bus.

The communication interface 570 supports wired/wireless Internet communication of the computing device 500 . The communication interface 570 may support various communication methods other than Internet communication. To this end, the communication interface 570 may be configured to include a communication module well known in the art.

The storage 590 may non-temporarily store one or more computer programs 591 . The storage 590 is a non-volatile memory such as a read only memory (ROM), an erasable programmable ROM (EPROM), an electrically erasable programmable ROM (EEPROM), a flash memory, a hard disk, a removable disk, or well in the art to which the present invention pertains. It may be configured to include any known computer-readable recording medium.

The computer program 591 may include one or more instructions in which methods/acts according to various embodiments described herein are implemented. When the computer program 591 is loaded into the memory 530 , the processor 510 may execute the one or more instructions to perform methods/operations according to various embodiments described herein.

The methods according to the embodiments described so far may be performed by executing a computer program implemented as computer readable code. The computer program may be transmitted from the first computing device to the second computing device through a network such as the Internet and installed in the second computing device, thereby being used in the second computing device. The first computing device and the second computing device include all of a server device, a physical server belonging to a server pool for cloud services, and a stationary computing device such as a desktop PC.

The computer program may be stored in a recording medium such as a DVD-ROM or a flash memory device.

Although the embodiments have been described above with reference to the accompanying drawings, those of ordinary skill in the art to which the present invention pertains can understand that the present invention can be embodied in other specific forms without changing the technical spirit or essential features. have. Therefore, it should be understood that the embodiments described above are illustrative in all respects and not restrictive.

Claims (17)

A method performed by a computing device, comprising:
constructing a database including a plurality of process recipes extracted from the thesis data;
selecting a first criterion from among a plurality of criteria for determining an optimal process from among the plurality of process recipes; and
determining the optimal process using the first criterion;
The step of constructing a database including a plurality of process recipes extracted from the thesis data includes:
extracting a first process sentence and a second process sentence appearing after the first process sentence from the thesis data;
identifying a process verb generating a target material from the first process sentence and extracting a subject and an object corresponding to the process verb; and
In the second process sentence, when the process verb is changed to an adjective and a subject and an object are not identified, determining the target material generated in the first process sentence as the subject of the second process sentence,
How to determine the optimal chemical process. According to claim 1,
The step of selecting a first criterion from among a plurality of criteria for determining an optimal process from among the plurality of process recipes,
Selecting a first criterion determined based on an amount of an intermediate produced in the process of performing each of the process recipes from among the plurality of criteria,
How to determine the optimal chemical process. 3. The method of claim 2,
Determining the optimal process using the first criterion comprises:
identifying intermediates generated in the process of performing each of the plurality of process recipes; and
Determining a process recipe having the smallest amount of the intermediate from among the plurality of process recipes as the optimal process,
How to determine the optimal chemical process. According to claim 1,
The step of selecting a first criterion from among a plurality of criteria for determining an optimal process from among the plurality of process recipes,
Including the step of selecting a first criterion according to the implementation difficulty of the environment in which the plurality of process recipes are performed,
How to determine the optimal chemical process. 5. The method of claim 4,
Determining the optimal process using the first criterion comprises:
determining a temperature for implementing a process to be performed with respect to a first process recipe among the plurality of process recipes; and
Comprising the step of determining the first process recipe as the optimal process when the temperature corresponds to a preset temperature,
How to determine the optimal chemical process. According to claim 1,
The step of selecting a first criterion from among a plurality of criteria for determining an optimal process from among the plurality of process recipes,
Including the step of selecting a first criterion according to the reliability of the process of performing the plurality of process recipes,
How to determine the optimal chemical process. 7. The method of claim 6,
Determining the optimal process using the first criterion comprises:
determining success rates of the plurality of process recipes by repeating the process of performing the plurality of process recipes; and
Comprising the step of determining the process recipe with the highest success rate among the plurality of process recipes as the process recipe with high reliability of the first criterion and determining the process recipe as the optimal process,
How to determine the optimal chemical process. According to claim 1,
The step of selecting a first criterion from among a plurality of criteria for determining an optimal process from among the plurality of process recipes,
Including the step of selecting a first criterion according to the cost of the material used in the process of performing the plurality of process recipes,
How to determine the optimal chemical process. 9. The method of claim 8,
Determining the optimal process using the first criterion comprises:
identifying a material used in the process of performing each of the plurality of process recipes; and
In the process performed for each of the plurality of process recipes, based on the total sum of the costs of the materials, determining a process recipe having the lowest total cost of the materials among the plurality of process recipes as the optimal process. ,
How to determine the optimal chemical process. According to claim 1,
Determining the optimal process using the first criterion comprises:
scoring each of the plurality of process recipes by using a score according to the weight assigned to the first criterion; and
Determining an optimal process according to the scoring result,
How to determine the optimal chemical process. 11. The method of claim 10,
Scoring each of the plurality of process recipes by using a score according to the weight assigned to the first criterion,
scoring each of the plurality of process recipes using a score obtained by adding a weight assigned to the first criterion and a weight assigned to a second criterion different from the first criterion among the plurality of criteria;
How to determine the optimal chemical process. 12. The method of claim 11,
The step of scoring each of the plurality of process recipes by using a score obtained by summing the weight assigned to the first criterion and the weight assigned to the second criterion,
excluding the process recipe from the optimal process when there is a process recipe having a score according to the weight assigned to the first criterion or a score according to the weight assigned to the second criterion is 0;
How to determine the optimal chemical process. According to claim 1,
Determining the optimal process using the first criterion comprises:
determining the optimal process by using a first process recipe selected based on the first criterion among the plurality of process recipes,
How to determine the optimal chemical process. delete According to claim 1,
The step of constructing a database including a plurality of process recipes extracted from the thesis data includes:
constructing a synonym list database for text having the same meaning by performing learning to analyze the meaning of the text included in the thesis data; and
extracting a process sentence from the thesis data using the synonym list database,
How to determine the optimal chemical process. processor;
network interface;
a memory executed by the processor to load a computer program; and
A computing device comprising a storage for storing the computer program,
The computer program is
instructions for constructing a database including a plurality of process recipes extracted from the thesis data;
instructions for selecting a first criterion from among a plurality of criteria for determining an optimal process from among the plurality of process recipes; and
instructions for determining the optimal process using the first criterion;
Instruction for building a database including a plurality of process recipes extracted from the thesis data,
instructions for extracting a first process sentence and a second process sentence appearing after the first process sentence from the thesis data;
instructions for identifying a process verb generating a target material from the first process sentence and extracting a subject and an object corresponding to the process verb; and
and instructions for determining the target material generated in the first process sentence as the subject of the second process sentence when the process verb is changed to an adjective in the second process sentence and a subject and an object are not identified.
computing device. A computer program for determining an optimal performance process comprising computer program instructions executable by a processor, wherein the computer program instructions are executed by a processor of a computing device, comprising:
constructing a database including a plurality of process recipes extracted from the thesis data;
selecting a first criterion from among a plurality of criteria for determining an optimal process from among the plurality of process recipes; and
a computer program for performing operations comprising determining the optimal process using the first criterion;
The step of constructing a database including a plurality of process recipes extracted from the thesis data includes:
extracting a first process sentence and a second process sentence appearing after the first process sentence from the thesis data;
identifying a process verb generating a target material from the first process sentence and extracting a subject and an object corresponding to the process verb; and
In the second process sentence, when the process verb is changed to an adjective and a subject and an object are not identified, determining the target material generated in the first process sentence as the subject of the second process sentence,
A computer program that performs operations is recorded,
computer readable recording medium.

Images