Design support device and design support method

Abstract

To automatically extract word phrases and numeric values related to components, functions and requirement specifications of an object product, and recognize and associate relations of the extracted word phrases and the extracted numeric values related to the components, the functions and the requirement specifications.SOLUTION: A design support device includes: an input part for acquiring design information related to an object product; an extraction part for extracting word phrases and numeric values related to components, functions and requirement specifications from the design information on the basis of classification axes stored in a classification axis storage part for storing classification axes being rules for classifying the word phrases and the numeric values related to the components, the functions and the requirement specifications in the object product; a recognition part for recognizing relations between the extracted word phrases and the extracted numeric values related to the components, the functions and the requirement specifications on the basis of a structure and co-occurrence of contents of the design information; an association part for associating the relations between the word phrases and the numeric values; and an output part for outputting results associated by the association part.SELECTED DRAWING: Figure 7

Description

[Technical field]

[0001] The present invention relates to a design support apparatus and a design support method. [Background technology]

[0002] In the field of product or information system (hereinafter referred to as "product") design, it is common to create new design proposals by modifying part of the design of an existing product. However, in recent years, products have become more complex and multifunctional, and there is a tendency for the number of parts and specifications to be considered when making design changes to increase. If the amount of work required for consideration becomes enormous, it becomes difficult to grasp the entire product, and product defects occur due to oversight.

[0003] In order to accurately grasp the impact of design changes, it is important to accurately understand the characteristics of the functions and required specifications that the design change part is responsible for. Also, if the interrelationships between the product's structure, functions, and required specifications are understood, it becomes possible to search for the parts of the design that should be changed based on the functions or required specifications that you want to change. Understanding the relationships between such structure, functions, and required specifications contributes to preventing design errors and improving design efficiency when design changes are made.

[0004] MBSE (Model-Based Systems Engineering) is used as a technology to get an overview of a product's structure, functions, and required specifications. MBSE aims to improve design efficiency and prevent design errors by visualizing the natural language used to describe the product as a model (diagram), and also visualizing the relationships between items included in each model. However, there is an issue that when linking parts, functions, and required specification items, the more items there are and the more complex the relationships, the more work required.

[0005] In relation to the above-mentioned problem, there is a technology for linking two pieces of natural language information, as disclosed in Patent Document 1. The design support device in Patent Document 1 extracts words related to the required specifications by executing natural language processing on the input required specification information, and extracts related technical information based on the extracted words. [Prior art documents] [Patent documents]

[0006] [Patent Document 1] Patent Publication No. 2021-131720 Summary of the Invention [Problem to be solved by the invention]

[0007] In order to grasp the relationships between terms and numerical values ​​related to parts, functions, and required specifications of a complex product, it is desirable to be able to acquire terms and numerical values ​​related to the parts, functions, and required specifications of the target product, and to automatically recognize the relationships between the acquired terms and numerical values ​​related to the parts, functions, and required specifications.

[0008] The design support device of Patent Document 1 does not extract items related to parts and functions from required specification information, and therefore is not sufficient as a means for extracting words and values ​​related to parts, functions, and required specifications of a product. The object of the present invention is to automatically extract words and numerical values ​​related to parts, functions, and required specifications of a target product, and to recognize and link the relationships between the words and numerical values ​​related to the extracted parts, functions, and required specifications. [Means for solving the problem]

[0009] The design support device of the present invention is characterized by comprising: an input unit for acquiring design information related to a target product; an extraction unit for extracting words and numerical values ​​related to the parts, functions, and required specifications from the design information based on a classification axis stored in a classification axis memory unit that stores a classification axis, which is a rule for classifying words and numerical values ​​related to the parts, functions, and required specifications of the target product; a recognition unit for recognizing relationships between the extracted words and numerical values ​​related to the parts, functions, and required specifications based on a structure and co-occurrence of the content of the design information; a linking unit for linking the relationships between the words and numerical values; and an output unit for outputting the linking results obtained by the linking unit. Other means will be described in the description of the embodiment of the invention. Effect of the Invention

[0010] According to the present invention, it is possible to automatically extract terms and values ​​related to parts, functions, and required specifications of a target product, recognize relationships between the terms and values ​​related to the extracted parts, functions, and required specifications, and perform linking. As a result, the present invention supports a bird's-eye view of a product by presenting the recognized relationships to a user. In addition, the present invention has various effects, such as supporting the prevention of omissions in inputting relationships between each term and numerical value when creating an MBSE model. [Brief description of the drawings]

[0011] [Figure 1] 1 is a block diagram of a design support device according to a first embodiment. [Figure 2A] FIG. 13 is a diagram for explaining classification axes. [Figure 2B] FIG. 13 is a diagram illustrating another example of the classification axis. [Diagram 3] FIG. 2 is a diagram illustrating an example of design information. [Figure 4] FIG. 13 is a diagram illustrating a processing procedure of an extraction unit. [Diagram 5] FIG. 13 is a diagram illustrating a processing procedure of a recognition unit. [Figure 6A] 13 is a flowchart of a recognition unit. [Figure 6B] 11 is another flowchart of the recognition unit. [Figure 7] 13 is an example of a result display screen. [Figure 8A] FIG. 11 is a block diagram of a design support device according to a second embodiment. [Figure 8B] FIG. 11 is another block diagram of the design support device according to the second embodiment. [Figure 9] 13 is a flowchart of a classification axis updating unit. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0012] Hereinafter, embodiments of the present invention will be described with reference to the drawings. Note that the drawings show specific embodiments according to the principles of the present invention, but these are for understanding the present invention and are in no way intended to limit the interpretation of the present invention.

[0013] Example 1 1 is a block diagram of a design support device 100 according to the first embodiment. The design support device 100 includes an input unit 101, an extraction unit 102, a recognition unit 103, a linking unit 104, an output unit 105, and a classification axis storage unit D001.

[0014] The input unit 101 acquires a plurality of pieces of design information related to a target product input by a user. The extraction unit 102 acquires words and numerical values ​​in the document of the design information acquired by the input unit 101 by natural language processing. The extraction unit 102 classifies and extracts words and numerical values ​​according to whether they correspond to a part, a function, or a required specification of the target product based on the classification axes stored in the classification axis storage unit D001. Natural language processing is a technique for mechanically processing language described in natural language to extract contents, and is used in search engines, predictive conversion, and the like. Note that the classification axis storage unit D001 does not have to be a component of the design support device 100 itself, and may be a component present in any location, such as on the cloud.

[0015] The recognition unit 103 recognizes the relationship between the classified and extracted words and numbers based on the structure and co-occurrence of the document content of the design information. Here, the structure of the document content refers to any structure that determines the dependency relationship between words and numbers, such as the relationship between subject and predicate, the relationship between modifier and modified word, the relationship between action and object of action, etc. Co-occurrence is the possibility (frequency) that two words and numbers appear at the same time. In other words, the recognition unit 103 recognizes the relationship between two words and numbers that frequently appear as a pair. Note that the "relationship between a word and a number" is a concept that includes the "relationship between a word and another word" and the "relationship between a number and another number."

[0016] The linking unit 104 links the relationships between the recognized words and the numerical values. Ultimately, the linking unit 104 links the relationships between the words and the numerical values ​​related to the parts, functions, and required specifications. The output unit 105 outputs the relationships between the words and numerical values ​​related to the parts, functions, and required specifications linked by the linking unit 104.

[0017] FIG. 2A is a diagram for explaining classification axes stored in the classification axis storage unit D001 according to the first embodiment. The classification axes according to the first embodiment are classification rules indicating whether a word or a numerical value is classified into a part, a function, or a required specification. Here, a part indicates the name of a part that constitutes a product or the name of a unit that is a collection of parts. A function indicates a characteristic of a product or a role of a product. A required specification indicates specific indicators related to characteristics and functions required for a product. The results of classification (parts, functions, and required specifications) are sometimes called "items."

[0018] 2B is a diagram illustrating another example of classification axes stored in the classification axis storage unit D001. The classification axis storage unit D001 may store words and numerical values ​​for each item of parts, functions, and required specifications.

[0019] Fig. 3 is a diagram showing an example of design information. Design information for product A is written in natural language. For example, the design information describes the functions of the parts that make up product A and the required specifications for performing those functions. There are two methods for describing design information: one is to structure the description content for each item of part, function, and required specifications (left diagram in Fig. 3), and the other is to describe without structuring (right diagram in Fig. 3).

[0020] 4 is a diagram showing a processing procedure of the extraction unit 102. The input unit 101 passes multiple pieces of design information acquired by itself to the extraction unit 102. The extraction unit 102 extracts words and numerical values ​​from documents included in the design information. To extract the words and numerical values, the extraction unit 102 breaks down the documents included in the design information into words from which the words and numerical values ​​can be extracted. To perform the decomposition process, the extraction unit 102 uses techniques such as a rule-based search method based on regular expressions, morphological analysis, and BERT (Bidirectional Encoder Representations from Transformers) in the field of natural language processing.

[0021] The extraction unit 102 breaks down the documents included in the design information into words, phrases, and numerical values, and then classifies each of the words and numerical values ​​based on the classification axes stored in the classification axis storage unit D001. The classification axis storage unit D001 stores classification axes that specify into which of the items of parts, functions, and required specifications the words and numerical values ​​are to be classified. The extraction unit 102 classifies the words and numerical values ​​based on the classification axes, and assigns the classification results (items) to the words and numerical values. When the design information is described structured for each item of parts, functions, and required specifications, the extraction unit 102 also extracts the items of parts, functions, and required specifications that are located at higher levels in the outline (hierarchical structure of paragraphs) of the structured document.

[0022] In addition, the extraction unit 102 may generate synonyms of the words and numbers stored in the classification axis storage unit D001, and classify the words and numbers in the design document including the synonyms, thereby performing more detailed classification. The classification axis storage unit D001 may store a dictionary in which synonyms have been learned.

[0023] Fig. 5 is a diagram showing the processing procedure of the recognition unit 103. The recognition unit 103 recognizes the relationships (described immediately below) between each word or phrase and numerical value in a document in which the extraction unit 102 has classified the words or phrases and numerical values. When the processing of Fig. 6A and Fig. 6B described below is completed, the recognition unit 103 generates the recognition result of Fig. 5.

[0024] FIG. 6A is a flowchart of the recognition unit 103. In step S601, the recognition unit 103 acquires the structure and co-occurrence of the contents of the document included in the design information in order to grasp the relationship between words and numerical values ​​that appear in the document. Here, the recognition unit 103 uses a structure analysis technique that grasps the relationship between a subject and a predicate, for example, in the field of natural language processing, and recognizes the co-occurrence of specific words and numerical values ​​with words and numerical values ​​that frequently appear in association with them.

[0025] In step S602, the recognition unit 103 acquires the words and numerical values ​​classified based on the classification axis by the extraction unit 102. Here, the recognition unit 103 acquires the words and numerical values ​​appearing in the documents included in the design information in order from the beginning of the documents included in the design information.

[0026] In step S603, the recognition unit 103 determines whether there are any words and numbers that may be related to the words and numbers obtained in step S602. The recognition unit 103 uses the structure and co-occurrence of the document obtained in step S601 to determine whether there is a "relationship" here. The words and numbers that may be related here are words and numbers that have a dependency relationship obtained by structural analysis, which is a technology in the field of natural language processing used by the recognition unit 103, or words and numbers that frequently co-occur in the entire document. The criterion for determining whether there is co-occurrence between words and numbers is whether the frequency of each word and number appearing in one sentence in the entire document is high. The recognition unit 103 may set a criterion for determining whether the frequency is high or not, and may be able to arbitrarily change the threshold value.

[0027] For example, if phrase B (predicate) exists for acquired phrase A (subject), the recognition unit 103 determines that "potentially related phrases and numerical values ​​exist." Note that the relationship between subject and predicate is an example of a dependency relationship. Furthermore, the relationship between a modifier and a modified word is also an example of a dependency relationship. If potentially related phrases and numerical values ​​exist (step S603 "Yes"), the recognition unit 103 proceeds to step S604, and otherwise (step S603 "No"), the recognition unit 103 proceeds to step S607.

[0028] In step S604, the recognition unit 103 determines whether the items (categories) of the potentially related words and numbers are of a different category from the items of the words and numbers acquired in step S602.

[0029] For example, as is clear from the bottom row of FIG. 5, if the acquired term "motor" is "part" and the potentially related term "connection" is "function", the recognition unit 103 determines that "motor" and "connection" are different items (classifications). On the other hand, if the acquired term "motor" is "part" and the potentially related term "gear" is also "part", the recognition unit 103 determines that "motor" and "gear" are the same item (classification). If the terms are different classifications (step S604 "Yes"), the recognition unit 103 proceeds to step S605, and otherwise (step S604 "No"), the recognition unit 103 proceeds to step S607.

[0030] In step S605, the recognition unit 103 determines whether or not the relationship between the words and numbers acquired in step S602 has already been recognized in a previous process, and acquires the result.

[0031] In step S606, the recognition unit 103 recognizes a final relationship between the words and numbers acquired in step S602 and the words and numbers that may be related. The recognition unit 103 does not need to recognize the relationship again for the numbers and words for which a relationship has already been recognized in step S605. As is clear from the bottom line of Fig. 5, the recognition unit 103 does not recognize a direct relationship between "motor" and "gear".

[0032] In step S607, the recognition unit 103 determines whether or not the process has been performed for all words, phrases, and numerical values. If the process has been performed (step S607 "Yes"), the recognition unit 103 ends the process related to the recognition of the relationship, and otherwise (step S607 "No"), the process returns to step S602.

[0033] Fig. 6B is another flowchart of the recognition unit 103. In Fig. 6A, the relationship between a word or phrase and a numerical value is recognized for each word or phrase and numerical value that appears in the design information. On the other hand, in Fig. 6B, the recognition unit 103 recognizes the relationship between all words or phrases and numerical values ​​included in the design information in one process. This shortens the processing time.

[0034] In step S608, the recognition unit 103 acquires the structure of the document included in the design information and the co-occurrence of words, phrases, and numerical values, similarly to step S601 in FIG. 6A. In step S609, the recognition unit 103 acquires all the words, phrases and numerical values ​​classified by the extraction unit 102 based on the classification axes. In step S610, the recognition unit 103 identifies related words and numbers for all words and numbers contained in the design information received by the input unit 101 based on the document structure and co-occurrence of words and numbers acquired in step S608. In step S611, the recognition unit 103 recognizes relationships between words and numbers in different categories among the related words and numbers identified in step S610.

[0035] The recognition unit 103 recognizes the relationship between the words and numerical values ​​related to the parts, functions, and required specifications from the entire design information by repeating the process. When the design information is described in a structured manner, the recognition unit 103 recognizes the relationship between the words and numerical values ​​related to the parts, functions, and required specifications included in the lower documents of the outline (hierarchical structure). Then, the recognition unit 103 recognizes the relationship between the words and numerical values ​​related to the parts, functions, and required specifications included in the higher documents of the outline. After that, the recognition unit 103 recognizes that there is also a relationship between the words and numerical values ​​related to the parts, functions, and required specifications included in the lower documents and the words and numerical values ​​related to the parts, functions, and required specifications included in the higher documents (see the second line in FIG. 5). However, the recognition unit 103 does not recognize the relationship across hierarchies between the words and numerical values ​​of the same classified items among the items of the parts, functions, and required specifications classified by the extraction unit 102.

[0036] Examples of higher-level and lower-level documents include basic design and detailed design, summary paragraph and specific examples, main text and appendix, main text and footnotes, an answer and multiple questions that elicit that answer, etc.

[0037] The linking unit 104 links the terms and numerical values ​​related to the parts, functions, and required specifications that have been recognized as being related by the recognition unit 103. For example, the linking unit 104 may assign identification numbers to the terms and numerical values ​​indicating the parts, functions, and required specifications extracted by the extraction unit 102, and assign to each of the terms and numerical values ​​the identification numbers of the terms and numerical values ​​of other items that have been recognized as being related by the recognition unit 103. In other words, the recognition unit 103 links a certain term and numerical value to the identification numbers of other terms and numerical values ​​that are the counterparts of the relationship.

[0038] In addition, in order to link the relationship recognized by the recognition unit 103 to which document of the design information it is described in, the linking unit 104 may assign an identification number to each document of the design information acquired by the input unit 101, and link the identification number to each word, phrase, and numerical value related to the corresponding parts, functions, and required specifications.

[0039] The linking unit 104 does not need to link all combinations of words that are recognized as being related. For example, the linking unit 104 may count the number of times that a word A and a word B co-occur in the same sentence in the design information, and if the number is less than a predetermined threshold, it may not be necessary to link the word A to the word B. As another example, the linking unit 104 may identify characteristic words of the design information by a known method (such as the tf-idf method) and link only combinations that include at least one characteristic word.

[0040] The output unit 105 displays the results of processing by the linking unit 104 (see FIG. 7).

[0041] 7 is an example of a result display screen 701. On the result display screen 701, the output unit 105 displays, for each item, the terms and numerical values ​​corresponding to the parts, functions, and required specifications extracted by the extraction unit 102, and also displays the relationships linked by the linking unit 104 with lines connecting the terms and numerical values.

[0042] The output unit 105 also displays in which design document the relationships between the terms and numerical values ​​related to the displayed parts, functions, and required specifications are described. For example, in the example shown in Fig. 7, when the user selects the lines indicating the relationships between "motor", "output", and "xxW", the output unit 105 displays the design information name "xx", the number of lines "xx", and the document "The motor output is xxW" in the "design information" column at the bottom of the result display screen 701. In addition to displaying the results, the output unit 105 may output an MBSE model indicating the relationships between the terms and numerical values ​​related to the parts, functions, and required specifications using the linked results.

[0043] Example 2 The design support device 100 in the second embodiment has a function of updating the classification axis storage unit D001 by using an MBSE model as an input, in addition to the functions of the design support device 100 in the first embodiment.

[0044] Fig. 8A is a block diagram of a design support device 100 according to Example 2. The design support device 100 in Fig. 8A includes an MBSE model input unit 801, an MBSE model information extraction unit 802, a synonym generation unit 803, a classification axis update unit 804, and a classification axis storage unit D001.

[0045] The MBSE model input unit 801 acquires a file describing an MBSE model from any device including an external device. The file describing the MBSE model is described in a modeling language such as SysML (Systems Modeling Language). The types of MBSE models are a block definition diagram describing only parts, a function block diagram describing only functions, and a requirement diagram describing only requirement specifications. In other words, each MBSE model includes only one of the items of parts, functions, and requirement specifications.

[0046] The MBSE model information extraction unit 802 extracts information from information described in a modeling language. That is, the MBSE model information extraction unit 802 extracts words and numerical values ​​describing components from a block definition diagram, words and numerical values ​​describing functions from a function block diagram, and words and numerical values ​​describing requirement specifications from a requirement diagram.

[0047] The synonym generating unit 803 generates synonyms for the words and numbers extracted by the MBSE model information extracting unit 802. The synonym generating unit 803 generates synonyms by utilizing a pre-learned database or the like. A synonym is another word and number having the same meaning as the original word and number. For example, for parts, "electric motor" is a synonym for "motor". For functions, "propulsion" is a synonym for "drive". For required specifications, "xx horsepower" is a synonym for the converted "◎W". This is also the case when the extracting unit 102 generates synonyms.

[0048] The classification axis update unit 804 updates the classification axes by storing the words and numerical values ​​obtained from the MBSE model and the generated synonyms as classification axes in the classification axis storage unit D001.

[0049] Fig. 8B is another block diagram of the design support device 100 according to the second embodiment. The configuration of the design support device 100 in Fig. 8B is the same as that shown in Fig. 8A except that the MBSE model input unit 801 is replaced with an MBSE model database D002. The MBSE model database D002 stores MBSE models of past cases related to the target product. The MBSE model database D002 updates the classification axis storage unit D001 by transferring the MBSE models of past cases from itself to the MBSE model information extraction unit 802.

[0050] FIG. 9 is a flowchart of the classification axis update unit 804. In step S901, the classification axis update unit 804 acquires the words and numerical values ​​acquired from the MBSE model information extraction unit 802, and the synonyms generated by the synonym generation unit 803. The acquired synonyms may be words or numerical values. In step S902, the classification axis update unit 804 classifies the words, numerical values, and synonyms acquired in step S901 into any one of parts, functions, and required specifications. The classification axis update unit 804 performs classification based on the type of the input MBSE model.

[0051] In step S903, the classification axis update unit 804 determines whether the classified words, numerical values, and synonyms are already stored as classification axes in the classification axis storage unit D001. If they are stored (step S903 "Yes"), the classification axis update unit 804 proceeds to step S906, and otherwise (step S903 "No"), proceeds to step S904. In step S904, the classification axis update unit 804 acquires the classification of the words and numerical values ​​and synonyms classified in step S902.

[0052] In step S905, the classification axis updating unit 804 stores (adds) the words, numerical values, and synonyms in the classification axis storage unit D001 in association with their classifications (items). In step S906, the classification axis update unit 804 determines whether the process has been performed for all of the words and numbers included in the input MBSE model and the generated synonyms. If the process has been performed (step S906 "Yes"), the classification axis update unit 804 ends the process, and otherwise (step S906 "No"), the classification axis update unit 804 returns to step S901. The classification axis update unit 804 performs the above-mentioned processing to update the classification axis using the input MBSE model, and classifies the words and numbers in the design information more accurately.

[0053] The effects of the design support system of this embodiment are as follows. (1) The design support device can visualize the relationship between parts, functions, and required specifications. (2) The computer-aided design system can generate synonyms of words and numbers from existing models written in a modeling language. (3) The computer-aided design system can accurately classify words and numbers through morphological analysis. (4) The design support system can recognize only the relationships that are truly necessary. (5) The design support device can also be applied to hierarchical design information. (6) The design support system can output a model of the relationships between words and numbers. (7) The design support device can extract synonyms of words and numbers that are specified in advance on the classification axis. (8) The design support device can recognize all relationships contained in a design document in a single process (high-speed processing).

[0054] The present invention is not limited to the above-described embodiment, but includes various modified examples. For example, the above-described embodiment has been described in detail to easily explain the present invention, and is not necessarily limited to those including all of the configurations described. It is also possible to replace a part of the configuration of one embodiment with the configuration of another embodiment, and it is also possible to add the configuration of another embodiment to the configuration of one embodiment. It is also possible to add, delete, or replace a part of the configuration of each embodiment with another configuration.

[0055] In addition, the above-mentioned configurations, functions, processing units, processing means, etc. may be realized in part or in whole by hardware, for example, by designing them as integrated circuits. In addition, the above-mentioned configurations, functions, etc. may be realized in software by a processor interpreting and executing a program that realizes each function. Information such as the program, table, file, etc. that realizes each function can be stored in a memory, a recording device such as a hard disk or SSD (Solid State Drive), or a recording medium such as an IC card, SD card, or DVD. In addition, the control lines and information lines are those considered necessary for the explanation, and not all control lines and information lines in the product are necessarily shown. In reality, it may be considered that almost all components are connected to each other. Various information of the inventory management device 1 may be stored on the cloud. [Explanation of symbols]

[0056] 100 Design support equipment 101 Input section 102 Extraction part 103 Recognition part 104 Stringing section 105 Output section 701 Result display screen 801 MBSE model input section 802 MBSE Model Information Extraction Unit 803 Thesaurus Generation Unit 804 Classification axis update section D001 Classification axis storage D002 MBSE model database

Claims

1. an input unit for acquiring design information regarding a target product; an extraction unit that extracts, from the design information, terms and values ​​related to the parts, functions, and required specifications based on a classification axis stored in a classification axis storage unit that stores a classification axis, the classification axis being a rule for classifying terms and values ​​related to the parts, functions, and required specifications of the target product; a recognition unit that recognizes relationships between extracted words and numerical values ​​related to the parts, functions, and required specifications based on a structure and co-occurrence of the contents of the design information; a linking unit for linking the relationship between the words and the numerical values; an output unit that outputs the result of the linking performed by the linking unit; A design support device comprising:

2. a thesaurus generating unit that generates, from a model describing the terms and values ​​related to the parts, functions, and required specifications of the target product, the thesaurus generating unit generating the thesaurus of the terms and values ​​related to the parts, functions, and required specifications described in the model; a classification axis updating unit that updates the classification axis storage unit by storing the words and numerical values ​​related to the parts, the functions, and the required specifications, and the generated synonyms, as the classification axes, in the classification axis storage unit; 2. The design support device according to claim 1, further comprising:

3. The extraction unit is Perform morphological analysis of natural language processing, classifying the words and numerical values ​​obtained by breaking down the document described in the design information into words based on the classification axis stored in the classification axis storage unit; 2. The design support device according to claim 1,

4. The recognition unit is Recognizing words and numbers that may be related to the words and numbers based on dependency relationships and co-occurrences of the words and numbers in the documents included in the design information; only when a classification of the words and numbers extracted by the extraction unit is different from a classification of the words and numbers that may be related, a relationship between the words and numbers extracted by the extraction unit and the words and numbers that may be related is recognized; 2. The design support device according to claim 1,

5. The recognition unit is When the design information is a structured document, recognizing the relationship between words and numbers located at higher levels of the hierarchy and words and numbers located at lower levels of the hierarchy; 2. The design support device according to claim 1,

6. The output unit is outputting a model showing terms and values ​​relating to the parts, functions and requirements, and relationships between the terms and values; 2. The design support device according to claim 1,

7. The extraction unit is generating synonyms related to the classification axis, and extracting words and numerical values ​​related to the parts, the functions, and the required specifications including the generated synonyms; 2. The design support device according to claim 1,

8. The recognition unit is Recognizing relationships between all words and values ​​contained in the design information in a single process; 2. The design support device according to claim 1,

9. The input unit of the design support device comprises: Obtain design information about the target product, The extraction unit of the design support device extracting, from the design information, terms and values ​​related to the parts, functions and required specifications based on a classification axis stored in a classification axis storage unit that stores a classification axis, which is a rule for classifying terms and values ​​related to the parts, functions and required specifications of the target product; The recognition unit of the design support device Recognizing relationships between the extracted words and values ​​related to the parts, functions, and requirements specifications based on the structure and co-occurrence of the content of the design information; The linking unit of the design support device Linking relationships between the words and numbers; The output unit of the design support device outputting a result of the linking performed by the linking unit; A design support method comprising:

Images