Model generation device and model generation method

The model generation device and method enhance drawing search efficiency by analyzing content parameters and calculating relevance scores, addressing the challenge of finding highly relevant drawings in conventional systems that rely on keyword-based searches.

JP7587111B2Active Publication Date: 2024-11-20NISSIN ELECTRIC CO LTD
View PDF 4 Cites 0 Cited by

Patent Information

Application Number
JP2020153966
Authority / Receiving Office
JP · JP
Patent Type
Patents
Current Assignee / Owner
Filing Date
2020-09-14
Publication Date
2024-11-20
Estimated Expiration
2040-09-14

AI Technical Summary

Technical Problem

Conventional drawing search systems, such as that described in Patent Document 1, do not efficiently allow users to quickly find desired drawings that are highly related to a target drawing, especially when keyword-based searches are challenging, and users may struggle to determine appropriate search terms.

Method used

A model generation device and method that generates a learning model to analyze drawing content parameters, calculate dissimilarity, and determine relevant drawings by focusing on specific character strings, enabling a relevance score-based ranking of search results without requiring keyword input.

Benefits of technology

Improves user convenience by efficiently identifying highly relevant drawings through a relevance score-based ranking, even when users cannot determine appropriate search terms, thus enhancing search accuracy and reducing the need to exhaustively check numerous results.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 0007587111000001
    Figure 0007587111000001
  • Figure 0007587111000002
    Figure 0007587111000002
  • Figure 0007587111000003
    Figure 0007587111000003
Patent Text Reader

Abstract

To provide a model generation device capable of improving user convenience in a drawing retrieval than before.SOLUTION: A learning device (11) calculates a difference degree in the description between any two different retrieval target drawings included in the multiple retrieval target drawings based on a content parameter of the retrieval target drawing in each of the multiple retrieval target drawings. The learning device (11) determines a learning non-target drawing out of the multiple retrieval target drawings according to the difference degree. The learning device (11) generates a learning model for retrieving at least one drawing corresponding to the target drawing out of the multiple retrieval drawings based on the content parameter of the retrieval target drawing for each of the multiple retrieval target drawings excluding the learning non-target drawing.SELECTED DRAWING: Figure 9
Need to check novelty before this filing date? Find Prior Art

Description

[Technical field]

[0001] One aspect of the present invention relates to a model generating device. [Background technology]

[0002] For example, in the field of plant engineering, it is necessary to handle a large number of drawings (e.g., past drawings). For this reason, there is a demand for technology to efficiently search for a desired drawing from a large number of drawings.

[0003] As an example, Patent Document 1 discloses a technology related to a user interface of a drawing search system (drawing management system). The drawing search system of Patent Document 1 accepts a selection operation of a device name and a drawing type name by a user, and searches for a group of drawings corresponding to the selection operation. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] JP 2003-99484 A Summary of the Invention [Problem to be solved by the invention]

[0005] According to the drawing search system of Patent Document 1, it is possible to present search results (group of drawings) corresponding to the above selection operation without requiring the user to input a search word. However, as will be described later, in the drawing search system of Patent Document 1, it is not necessarily easy for the user to quickly find a desired drawing corresponding to the target drawing (e.g., a specific drawing considered to be highly related to the target drawing) from the search results.

[0006] As described above, the conventional drawing search technology has room for improvement in terms of improving user convenience. One aspect of the present invention aims to improve user convenience in drawing searches more than ever before. [Means for solving the problem]

[0007] In order to solve the above problems, a model generation device according to one embodiment of the present invention is a model generation device that generates a learning model for searching for at least one drawing corresponding to a target drawing from a plurality of search target drawings, and includes: a search target drawing content parameter acquisition unit that acquires search target drawing content parameters related to the description content of each of the plurality of search target drawings by analyzing the plurality of search target drawings; a dissimilarity calculation unit that calculates the difference in the description content between any two different search target drawings included in the plurality of search target drawings based on the search target drawing content parameters of each of the plurality of search target drawings; a determination unit that determines drawings that are not to be learned from among the plurality of search target drawings in accordance with the difference; and a learning unit that generates the learning model based on the search target drawing content parameters of each of the plurality of search target drawings excluding the drawings that are not to be learned.

[0008] In addition, in order to solve the above problem, a model generation method according to one embodiment of the present invention is a model generation method for generating a learning model for searching for at least one drawing corresponding to a target drawing from a plurality of search target drawings, and includes a search target drawing content parameter acquisition step of acquiring search target drawing content parameters related to the description contents of each of the plurality of search target drawings by analyzing the plurality of search target drawings, a difference calculation step of calculating the difference in the description contents between any two different search target drawings included in the plurality of search target drawings based on the search target drawing content parameters of each of the plurality of search target drawings, a determination step of determining drawings that are not to be learned from among the plurality of search target drawings based on the difference, and a learning step of generating the learning model based on the search target drawing content parameters of each of the plurality of search target drawings excluding the drawings that are not to be learned. Effect of the Invention

[0009] According to one aspect of the present invention, it is possible to improve user convenience in drawing searches compared to the conventional art. [Brief description of the drawings]

[0010] [Figure 1] FIG. 1 is a block diagram showing a configuration of a main part of an information processing system according to a reference embodiment. [Diagram 2] FIG. 11 is a diagram showing an example of a first table in the reference embodiment. [Diagram 3] 11A and 11B are diagrams illustrating an example of a specific character string extraction process for a past drawing. [Figure 4] FIG. 11 is a diagram showing an example of a second table in the reference embodiment. [Diagram 5] FIG. 13 is a diagram showing an example of a third table in the reference embodiment. [Figure 6] 11A and 11B are diagrams illustrating an example of a specific character string extraction process for a new drawing. [Figure 7] 13A and 13B are diagrams illustrating an example of deriving a relevance score in a reference embodiment. [Figure 8] FIG. 13 is a diagram showing an example of a drawing search result in the reference embodiment. [Figure 9] 1 is a block diagram showing a configuration of a main part of an information processing system according to a first embodiment. [Figure 10] FIG. 4 is a diagram illustrating a difference degree in the first embodiment. [Figure 11] FIG. 4 is a diagram showing an example of a dissimilarity table in the first embodiment. [Figure 12] FIG. 4 is a diagram illustrating an example of a process of a similar set determination unit in the first embodiment. [Figure 13] FIG. 2 is a diagram showing an example of a dendrogram representing the hierarchical structure of a set obtained in the first embodiment. [Figure 14] 11 is a diagram illustrating an example of processing by a non-learning drawing determination unit in the first embodiment. FIG. [Figure 15] FIG. 11 is a diagram showing an example of an edited second table in the first embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0011] [Reference form] Prior to the description of the information processing system 100 of the first embodiment, an information processing system 100s will be described as a reference embodiment. For convenience of description, in each of the following embodiments, the same reference numerals are given to components having the same functions as those described in the reference embodiment, and the description thereof will not be repeated. In addition, for the sake of brevity, the description of matters similar to those in the publicly known technology will be omitted as appropriate.

[0012] Please note that each numerical value described below in this specification is merely an example. In this specification, the description "A to B" for two numbers A and B means "greater than or equal to A and less than or equal to B" unless otherwise specified.

[0013] (Overview of Information Processing System 100s) 1 is a block diagram showing the configuration of a main part of an information processing system 100s. The information processing system 100s includes an information processing device 1s, a past property plan DB (Database) 91, and a new property plan DB 92.

[0014] The information processing device 1s includes a control device 10s, an input unit 71, a display unit 72, and a storage unit 80. The control device 10s includes a learning device 11s and a drawing search device 12. In this specification, the term "property" refers to, for example, a "site" in plant engineering.

[0015] The information processing device 1s may be communicatively connected to the past property plan DB 91 and the new property plan DB 92. Therefore, unlike the example of FIG. 1, at least one of the past property plan DB 91 and the new property plan DB 92 may be provided inside the information processing device 1s.

[0016] Drawings to be searched for by the information processing device 1s (more specifically, the control device 10s) (search target drawings) include, for example, specification drawings, design drawings, and production drawings. Search target drawings may also include specifications, design documents, and estimates. In this way, search target drawings are not limited to documents containing the name "drawings." As an example, search target drawings include any type of document related to project plans in the plant engineering field.

[0017] The control device 10s comprehensively controls each unit of the information processing device 1s. The storage unit 80 stores various data and programs used in the processing of the control device 10s. As described below, the control device 10s uses machine learning to search for at least one drawing corresponding to the target drawing (e.g., drawing ND) from a plurality of search target drawings (e.g., past drawings a1 to MN).

[0018] The input unit 71 accepts user operations (user operations). The display unit 72 displays various data. As an example, the display unit 72 may display data indicating a search result by the control device 10s. The input unit 71 and the display unit 72 may be provided as one unit. For example, the input unit 71 and the display unit 72 can be integrated by using a touch panel.

[0019] (Previous property drawing DB91) The past property drawing DB91 stores drawings (strictly speaking, drawing data) relating to each past property (each existing property). In the following explanation, "drawing data of drawing A (a certain drawing)" will be abbreviated simply to "drawing A" as appropriate. Also, "drawing number of drawing A" will be abbreviated simply to "drawing A" as appropriate.

[0020] In a reference embodiment, the past property drawing DB 91 stores drawings for each of a plurality of M different properties. M is an integer equal to or greater than 1. Hereinafter, the jth property is also referred to as "property j." j is an integer equal to or greater than 1 and equal to or less than M.

[0021] Additionally, the past property drawing DB91 stores N different drawings (N types) for each of properties 1 to M. N is an integer equal to or greater than 1. Hereinafter, the i-th drawing for property j will also be referred to as "drawing (i,j)." Additionally, the i-th drawing (i-th type) for each property will also be collectively referred to as drawing i. i is an integer equal to or greater than 1 and equal to or less than N.

[0022] As described above, a total of T drawings are stored in the past property drawing DB 91. In the example of the reference embodiment, T=M×N. T is an integer of 2 or more. In other words, at least one of M and N is 1 or more.

[0023] Specifically, in the past property drawing DB91, each type of drawing number is listed by property number in the form of the first table TB1 shown in Figure 2. Hereinafter, the first table TB1 will also be abbreviated to "TB1". Other elements will also be abbreviated in the same manner as appropriate. TB1 is also called the "past property-drawing correspondence table sorted by property". The cell in the i-th row and j-th column of TB1 indicates the drawing number of drawing (i,j).

[0024] In the example of Fig. 2, for convenience, properties 1 to 3 are also written as properties A to C, respectively. In the example of Fig. 2, the first type of drawing (Drawing 1) is an outline drawing, the second type of drawing (Drawing 2) is an assembly drawing, the third type of drawing (Drawing 3) is a foundation drawing, and the Nth type of drawing (Drawing N) is a configuration drawing.

[0025] For simplicity, drawings (1,1) to (N,1) in the example of Fig. 2 (i.e., N types of drawings for property A) will be referred to as drawings a1 to aN below. For example, drawings a1 to aN in the example of Fig. 2 respectively refer to the outline drawing to the configuration drawing of property A. The drawings for the other properties will be referred to in the same manner.

[0026] As described above, the past property drawing DB 91 stores a total of T drawings, from drawing a1 to drawing MN. Hereinafter, drawings a1 to MN are also collectively referred to as past drawings. Past drawings are an example of drawings to be searched. For this reason, drawings a1 to MN are also referred to as a group of drawings to be searched.

[0027] In addition, in this specification, any one of the multiple past drawings (search target drawings) is also referred to as a candidate drawing. As an example, the processing for drawing a1 (exterior drawing of property A) as a candidate drawing will be mainly illustrated. The processing for other drawings will be omitted as appropriate, but is the same as the case of drawing a1.

[0028] (Learning device 11s) The learning device 11s includes a past drawing data acquisition unit 111, a past drawing data preprocessing unit 112 (candidate drawing content parameter acquisition unit, search target drawing content parameter acquisition unit), and a learning model generation unit 113s. The learning device 11s generates a learning model for drawing search by the drawing search device 12 based on the drawings a1 to MN. For this reason, the learning device 11s may be referred to as a model generation device. An example of the processing flow of the learning device 11s will be described below.

[0029] (Obtaining candidate drawings) In the past property drawing DB91, drawings a1 to MN are pre-sorted by drawing type according to TB1s. Therefore, for example, the past drawing data acquisition unit 111 acquires each exterior drawing from the past property drawing DB91 in the order of "drawing a1 → b1 → ... → M1". Next, the past drawing data acquisition unit 111 acquires each assembly drawing from the past property drawing DB91 in the order of "drawing a2 → b2 → ... → M2". And finally, the past drawing data acquisition unit 111 acquires each configuration drawing from the past property drawing DB91 in the order of "drawing aN → bN → ... → MN".

[0030] In the above example, the past drawing data acquisition unit 111 first refers to the cell in the first row and first column of TB1s. Then, the past drawing data acquisition unit 111 acquires the drawing (1,1) corresponding to the above cell, that is, drawing a1, from the past property drawing DB 91. The past drawing data acquisition unit 111 supplies the acquired drawing a1 to the past drawing data preprocessing unit 112.

[0031] (Generating pre-processing data corresponding to candidate drawings) The past drawing data pre-processing unit 112 analyzes the drawing a1 to generate pre-processed data corresponding to the drawing a1. As an example, the past drawing data pre-processing unit 112 performs OCR processing on the drawing a1 to generate post-OCR processing data. In the post-OCR processing data, each character represented as image data in the drawing a1 is converted into a character code.

[0032] Then, the past drawing data preprocessing unit 112 extracts (detects) a specific character string that is preset for each drawing from the OCR processed data. Hereinafter, the specific character string is referred to as a specific character string. In the first embodiment, it is assumed that L (L is an integer equal to or greater than 2) different k-th specific character strings are preset.

[0033] In the following, the kth specific string is referred to as the kth specific string, where k is an integer equal to or greater than 1 and equal to or less than L. In the following description, a case will be illustrated in which the first specific string is set to a "voltage value", the second specific string is set to a "current value", the third specific string is set to "OR", and the Lth specific string is set to "open".

[0034] However, the specific character string is not limited to the above example. The specific character string may be, for example, a word that is commonly used to explain product specification information (technical information) in drawings in the field of engineering.

[0035] 3 is a diagram showing an example of a specific character string extraction process for drawing a1. In the example of FIG. 3, a character string "voltage value 90V" is written in an area L1 of drawing a1. In this case, the past drawing data pre-processing unit 112 extracts a series of character strings "voltage value 90V" in L1 from the data after OCR processing. The past drawing data pre-processing unit 112 further extracts a first specific character string "voltage value" from the extracted series of character strings.

[0036] Next, the past drawing data pre-processing unit 112 applies a known syntax analysis technique to the series of character strings, and determines that the character string "90V" following the first specific character string "Voltage value" indicates the attribute of "Voltage value". In this case, the past drawing data pre-processing unit 112 acquires the number "90" contained in the character string "90V" as a value indicating the magnitude of the "Voltage value". In this way, the past drawing data pre-processing unit 112 determines by syntax analysis that "the series of character strings described in L1 indicate the specification information that 'the voltage value is 90V.'"

[0037] Moreover, a series of character strings "current value 1.0A" is described in the area L2 of the drawing a1. The past drawing data pre-processing unit 112 extracts the second specific character string "current value" from the series of character strings in the same manner as the example of L1. The past drawing data pre-processing unit 112 then determines that the character string "1.0A" following the second specific character string "current value" indicates the attribute of "current value". Therefore, the past drawing data pre-processing unit 112 acquires the number "1.0" contained in the character string "1.0A" as a value indicating the magnitude of the "current value". In this way, the past drawing data pre-processing unit 112 determines that "the series of character strings described in L2 indicate the specification information that 'the current value is 2.0A'".

[0038] However, the drawing a1 includes a third specific character string "OR", but the third specific character string "OR" is not followed by any character string. In this case, the past drawing data pre-processing unit 112 detects the number of occurrences of the third specific character string "OR" included in the OCR processed data. As a result of the detection process, the past drawing data pre-processing unit 112 determines that the number of occurrences of the third specific character string "OR" included in the drawing a1 is three (see area L3).

[0039] Similarly, the drawing a1 includes the Lth specific character string "Open", but no characters follow the Lth specific character string "Open". Therefore, the past drawing data pre-processing unit 112 determines that the number of the Lth specific character string "Open" included in the drawing a1 is one, as in the example of L3 (see area L4).

[0040] Then, the previous drawing data preprocessing unit 112 generates preprocessed data corresponding to the drawing a1 based on the above-mentioned analysis result of the OCR processed data. The preprocessed data is a data set indicating the k-th content parameter (hereinafter, Ak) associated with the k-th specific character string. The first to L-th content parameters are also collectively referred to as content parameters.

[0041] It can be said that the content parameter is one of the data that quantifies (quantifies) the contents of the description in the drawing (specifically, the contents related to the specific character string). For this reason, the content parameter is used as an index that indicates the contents of the description in the drawing. Note that, to distinguish from the target drawing content parameters described later, the content parameters of the search target drawing (past drawing) are also referred to as the search target drawing content parameters. Also, the kth content parameter of the search target drawing is also referred to as the search target drawing kth content parameter.

[0042] In the following, Ak of ​​drawing (i, j) is also written as Ak(i, j). As described above, the past drawing data preprocessing unit 112 sets Ak(i, j) based on the analysis result for drawing (i, j) (more specifically, the detection result of a specific character string in drawing (i, j)).

[0043] As an example, the past drawing data pre-processing unit 112 sets A1=90 for drawing a1 (see also the hatched portion in FIG. 4 described later). That is, the past drawing data pre-processing unit 112 sets the value "90" indicating the magnitude of the first specific character string "voltage value" written on drawing a1 as the first content parameter.

[0044] Similarly, the past drawing data pre-processing unit 112 sets A2=1.0. That is, the past drawing data pre-processing unit 112 sets the value "1.0" indicating the magnitude of the second specific character string "current value" described in the drawing a1 as the second content parameter.

[0045] In response to this, the past drawing data pre-processing unit 112 sets A3=3. That is, the past drawing data pre-processing unit 112 sets the value "3" indicating the number of times the third specific character string "OR" appears in the drawing a1 as the third content parameter.

[0046] Similarly, the past drawing data pre-processing unit 112 sets AL=1. That is, the past drawing data pre-processing unit 112 sets the value "1" indicating the number of times the Lth specific character string "Open" is written in the drawing a1 as the Lth content parameter.

[0047] As described above, the past drawing data pre-processing unit 112 sets A1 to AL for the drawing a1. As described above, the past drawing data pre-processing unit 112 acquires the content parameters of the candidate drawing by analyzing the candidate drawing. For this reason, the past drawing data pre-processing unit 112 is also called a candidate drawing content parameter acquisition unit.

[0048] The past drawing data pre-processing unit 112 also performs similar processing on the other past drawings. That is, the past drawing data pre-processing unit 112 sets A1 to AL for the drawings a1 to MN, respectively.

[0049] Thereafter, the previous drawing data pre-processing unit 112 generates a second table TB2 indicating A1 to AL of each of the drawings a1 to MN, i.e., A1(1,1) to AL(M,N). As described below, TB2 is an example of teacher data in the reference embodiment. That is, the previous drawing data pre-processing unit 112 also operates as a functional unit (teacher data generating unit) that generates teacher data.

[0050] TB2 includes an ith sub-table TB2-i in the second table. TB2-i is a table showing A1 to AL of each of the drawings i. FIG. 4 shows TB2-1 as an example. TB2-1 shows A1 to AL of each of the drawings a1 to M1 (exterior drawings of properties A to M). TB2s is composed of a set of N sub-tables, TB2-1 to TB2-N. In this way, in the reference embodiment, a sub-table is created for each drawing type.

[0051] (Generation of learning models in reference form) The learning model generation unit 113s generates a learning model according to the drawings a1 to M1 based on pre-processing data corresponding to the drawings a1 to M1 (i.e., A1 to AL in the drawings a1 to M1). A known machine learning algorithm may be used to generate the learning model. As an example, in the first embodiment, the learning model generation unit 113s generates the learning model by multinomial logistic regression analysis.

[0052] As described above, TB2 indicates the number (an example of an identifier) ​​of each of the multiple past drawings (e.g., drawing a1) and the corresponding relationship between each of the multiple past drawings A1 to AL. Therefore, the learning model generation unit 113s acquires TB2 as training data. Then, the learning model generation unit 113s generates a learning model in the reference form by performing a multinomial logistic regression analysis using the training data.

[0053] As an example, the learning model generation unit 113s calculates the k-th weighting coefficient (hereinafter, Bk) associated with the k-th specific character string in the drawing a1 based on each of A1 to AL of TB2-1. The first to L-th weighting coefficients are also collectively referred to as weighting coefficients. Moreover, Bk in the drawing (i,j) is also written as Bk(i,j). Bk(i,j) is a weighting coefficient corresponding to Ak(i,j). The weighting coefficient is used to calculate the relevance score, which will be described later.

[0054] For example, the learning model generation unit 113s calculates B1 to BL in the drawing a1 by the above-mentioned multinomial logistic regression analysis. In the reference embodiment, for the drawing a1, the weighting coefficients are calculated as B1=-0.5, B2=-1.2, B3=0.6, ..., BL=0.7 (see the hatched areas in FIG. 5 described later).

[0055] The learning model generation unit 113s performs similar processing for other past drawings. That is, the learning model generation unit 113s sets B1 to BL for each of the drawings a1 to MN. In this way, the learning model generation unit 113s generates a learning model including B1 to BL for each of the drawings a1 to MN, i.e., B1(1,1) to BL(M,N).

[0056] Then, the learning model generation unit 113s generates a third table TB3 indicating B1(1,1) to BL(M,N). TB3 includes an i-th sub-table TB3-i in the third table. TB3-i is a table indicating B1 to BL of each of the drawings i. FIG. 5 shows TB3-1 as an example. TB3-1 shows B1 to BL of each of the drawings a1 to M1. In this way, TB3-1 indicates a weighting coefficient corresponding to the content parameter of TB2-1. TB3 is composed of a set of N sub-tables, TB3-1 to TB3-N.

[0057] (New property drawing DB92) In the new property drawing DB 92, drawings (hereinafter collectively referred to as new drawings) relating to new properties (e.g., at least one property that is scheduled to be constructed) are stored as a new property data set. In the following, one new property (property T) included in the new property data set will be described.

[0058] As an example, the new property drawing DB92 stores N different drawings (from exterior drawing to configuration drawing) of the same type as those of past properties for property T. In the following, the exterior drawing of property T is illustrated as an example of a new drawing. The exterior drawing of property T is also referred to as drawing ND (see also FIG. 6 described later).

[0059] (Drawing search device 12) The drawing search device 12 includes a new drawing data acquisition unit 121, a new drawing data preprocessing unit 122 (target drawing content parameter acquisition unit), a score calculation unit 123, and a search result data generation unit 124. The score calculation unit 123 and the search result data generation unit 124 may be collectively referred to as a search unit.

[0060] In the reference embodiment, the drawing search device 12 searches for at least one drawing by matching the target drawing with each of a plurality of search target drawings using a learning model generated by the learning device 11s. The drawing ND in the reference embodiment is an example of a target drawing. As described below, the drawing search device 12 uses the learning model to match the drawing ND with each of the drawings a1 to MN.

[0061] (Getting new drawings) The new drawing data acquisition unit 121 is a functional unit paired with the past drawing data acquisition unit 111. As an example, the new drawing data acquisition unit 121 acquires a predetermined new drawing (e.g., drawing ND) included in the new property data set of the new property drawing DB 92, when the input unit 71 accepts a predetermined user operation. The new drawing data acquisition unit 121 supplies the acquired drawing ND to the new drawing data pre-processing unit 122.

[0062] (Generating pre-processing data for new drawings) The new drawing data pre-processing unit 122 is a functional unit paired with the previous drawing data pre-processing unit 112. The new drawing data pre-processing unit 122 generates pre-processed data corresponding to the drawing ND by the same process as the previous drawing data pre-processing unit 112. That is, the new drawing data pre-processing unit 122 performs OCR processing on the drawing ND and extracts a specific character string from the drawing ND. Then, the new drawing data pre-processing unit 122 sets content parameters of the drawing ND based on the extraction result of the specific character string. Hereinafter, the k-th content parameter of the drawing ND is also referred to as Ck.

[0063] That is, the new drawing data pre-processing unit 122 acquires the content parameters of the drawing ND by analyzing the drawing ND using the same analysis method as the past drawing data pre-processing unit 112. Note that, in order to distinguish from the above-mentioned search target drawing content parameters, the content parameters of the target drawing (drawing ND) are also referred to as target drawing content parameters. Also, the k-th content parameter of the target drawing is also referred to as target drawing k-th content parameter.

[0064] Fig. 6 is a diagram showing an example of a specific character string extraction process for a drawing ND. Fig. 6 is a diagram paired with Fig. 3. In the example of Fig. 6, a character string "Voltage value 50V" is written in an area R1 of the drawing ND. Therefore, the new drawing data pre-processing unit 122 sets C1=50 based on the extraction result of the first specific character string (see also Fig. 7 described later).

[0065] Incidentally, the position of region R1 in drawing ND is different from the position of region L1 in drawing a1 (see Figs. 3 to 6). However, although the description position in drawing ND is different from that in drawing a1, the same type of specification information (voltage value) is described as in drawing a1. According to the method of setting the first content parameter by the new drawing data pre-processing unit 122, the agreement of the description contents (specification information related to the first specific character string) between drawing ND and drawing a1 can be quantitatively evaluated regardless of the apparent difference between drawing ND and drawing a1 (e.g., the difference in the description position of the first specific character string). In this regard, the same applies to C2 to CL (in other words, regions R2 to R4) described below.

[0066] 6, the character string "current value 2.5A" is written in the region R2 of the drawing ND. Therefore, the new drawing data pre-processing unit 122 sets C2=2.5 based on the extraction result of the second specific character string (see FIG. 6).

[0067] Furthermore, the new drawing data pre-processing unit 122 determines that the number of occurrences of the third specific character string "OR" included in the drawing ND is three (see region R3 in FIG. 6). Therefore, the new drawing data pre-processing unit 122 sets C3=3.

[0068] Similarly, the new drawing data pre-processing unit 122 determines that the number of occurrences of the Lth specific character string "KAI" included in the drawing ND is two (see region R4 in FIG. 6). Therefore, the new drawing data pre-processing unit 122 sets CL=2.

[0069] As described above, the new drawing data pre-processing unit 122 sets C1 to CL by performing the same processing on the drawing ND as the past drawing data pre-processing unit 112. After that, the new drawing data pre-processing unit 122 generates a new drawing content parameter table TB-ND indicating C1 to CL (see FIG. 7).

[0070] (Calculation of relevance score) The score calculation unit 123 calculates a relevance score based on the learning model (more specifically, B1(1,1) to BL(M,N)) generated by the learning model generation unit 113s. The relevance score is a score (index) indicating the degree of relevance of each past drawing (drawings a1 to MN) to drawing ND. Hereinafter, the relevance score of drawing (i,j) is also represented as SCORE(i,j). SCORE(i,j) is also abbreviated as SCORE.

[0071] In the following, a case where a relevance score is calculated will be illustrated with reference to FIG. 7 (particularly the hatched portion in FIG. 7). In the following example, for convenience of explanation, a case where a relevance score is calculated for each of past drawings of the same type as drawing ND (i.e., drawings a1 to M1 which are outline drawings) will be described. However, it should be noted that for past drawings of a type different from drawing ND, relevance scores are derived in the same manner as in the example of FIG. 7.

[0072] Fig. 7 is a diagram for explaining an example of deriving a relevance score. Fig. 7 mainly describes an example of calculating the SCORE of drawing a1. The score calculation unit 123 calculates the first raw score to the Lth raw score of the drawing a1 using (i) C1 to CL derived by the new drawing data preprocessing unit 122 and (ii) B1 to BL of the drawing a1 derived by the learning model generation unit 113s.

[0073] Hereinafter, the kth raw score is also written as SSk. The kth raw score of drawing (i,j) is also written as SSk(i,j). The first raw score to the Lth raw score are also collectively called raw scores. SSk corresponds to Bk (the kth weighting coefficient).

[0074] Specifically, the score calculation unit 123 calculates the product of the k-th content parameter (Ck) of the drawing ND and the k-th weighting coefficient (Bk) of the drawing a1 as SSk. SSk = Ck × Bk … (1) Calculate.

[0075] In the example of Fig. 7, C1 = 50, B1 = -0.5. Therefore, the score calculation unit 123 derives SS1 as SS1 = 50 x -0.5 = -25. In the same manner, the score calculation unit 123 calculates each subsequent raw score. For example, the score calculation unit 123 calculates each raw score as SS2 = -3, SS3 = 1.8, ..., SSL = 1.4.

[0076] Next, the score calculation unit 123 calculates a relevance score (SCORE) of the drawing a1 based on the first raw score to the Lth raw score of the drawing a1. Specifically, the score calculation unit 123 calculates the sum of each raw score from the first raw score to the Lth raw score as the SCORE. That is, the score calculation unit 123 calculates the relevance score (SCORE) of the drawing a1 based on the first raw score to the Lth raw score of the drawing a1. SCORE = ΣSSk … (2) The range of k on the right side of formula (2) is from 1 to L. This also applies to formula (2A) etc. described later.

[0077] As described above, the score calculation unit 123 calculates the SCORE using each Ck and each Bk. Hereinafter, the relevance score of the drawing a1 is also written as SCORE(a1). The relevance scores of the other past drawings are also written in the same manner. In the example of FIG. 7, the result of the calculation by the score calculation unit 123 is SCORE(a1)=-30.4.

[0078] Next, the score calculation unit 123 calculates the SCORE of each of the remaining past drawings (drawings b1 to M1 in the example of FIG. 7) in a similar manner. In the example of FIG. 7, the calculation results obtained by the score calculation unit 123 are SCORE(b1)=41.1, SCORE(c1)=36.5, ..., SCORE(M1)=4.2.

[0079] In the example of Figure 7, it was confirmed that the maximum relevance score of the M scores, SCORE(a1) to SCORE(M1), was SCORE(b1). It was also confirmed that SCORE(c1) was the second maximum of the M scores. It was also confirmed that the minimum relevance score of the M scores was SCORE(a1).

[0080] The score calculation unit 123 generates a fourth table TB4 indicating its own calculation results. TB4 includes an i-th sub-table TB4-i in the fourth table. TB4-i is a table indicating SS1 to SSL and SCORE of each of the drawings i. FIG. 7 shows TB4-1 as an example. TB4-1 shows SS1 to SSL and SCORE of each of the drawings a1 to M1. In this way, TB4-1 is generated based on TB-ND and TB3-1. TB4 is composed of a set of N sub-tables, TB4-1 to TB4-N.

[0081] In addition, the above formula (1) is SSk(i,j)=Ck×Bk(i,j) …(1A) In this way, the score calculation unit 123 calculates SSk(i,j) based on Bk(i,j).

[0082] Moreover, the above formula (2) is SCORE(i,j) = ΣSSk(i,j) =Σ{Ck×Bk(i,j)}…(2A) In this way, the score calculation unit 123 calculates SCORE(i,j) based on each SSk(i,j) (that is, each Ck and each Bk(i,j)).

[0083] Moreover, the above formula (2A) is SCORE(i,j)=Σ{Ak(i,j)×Bk(i,j)}…(3) It can also be understood as an equation in which Ck is substituted for Ak(i,j) in the above equation. In this way, the score calculation unit 123 can calculate the SCORE, which is a dependent variable (objective variable), by applying C1 to CL as independent variables (explanatory variables) in the learning model. The right-hand side of equation (3) is an example of a linear model in multinomial logistic regression analysis (multinomial logistic regression analysis model).

[0084] In addition, SSk can also be expressed as the output of the learning model according to Ck. From this, it can also be expressed that the score calculation unit 123 calculates the SCORE based on the outputs (SS1 to SSk) of the learning model.

[0085] Alternatively, the learning model may be configured to output a SCORE. In this case, the score calculation unit 123 obtains the SCORE as an output of the learning model, and supplies the SCORE to the search result data generation unit 124.

[0086] (Decision of the order of candidates) The search result data generating unit 124 determines a candidate order (candidate ranking) corresponding to the drawing ND based on the SCORE(a1) to SCORE(M1) calculated by the score calculating unit 123. Specifically, the search result data generating unit 124 determines a candidate order (first candidate to Mth candidate) for each of the drawings a1 to M1 (outline drawings) based on the SCORE(a1) to SCORE(M1).

[0087] The search result data generation unit 124 sorts SCORE(a1) to SCORE(M1) according to the value of each SCORE. As an example, the search result data generation unit 124 sorts SCORE(a1) to SCORE(M1) in descending order of value. Then, the search result data generation unit 124 determines the past drawing corresponding to the gth largest SCORE as the gth candidate (hereinafter also referred to as candidate g). g is an integer greater than or equal to 1 and less than or equal to M.

[0088] By setting the candidate number according to the SCORE in this way, it is expected that the smaller the candidate number (the higher the candidate), the higher the relevance between the past drawing and the drawing ND. For example, candidate 1 refers to the past drawing that is shown to have the highest relevance to the drawing ND among the drawings a1 to M1 according to the SCORE.

[0089] As described above, SCORE(b1) is the maximum relevance score (the largest relevance score) among SCORE(a1) to SCORE(M1). Therefore, the search result data generation unit 124 determines drawing b1 as candidate 1 (see the hatched area in FIG. 8). In this way, the search result data generation unit 124 selects drawing b1 as the past drawing most highly related to drawing ND.

[0090] Moreover, SCORE(c2) is the second largest relevance score among SCORE(a1) to SCORE(M1). Therefore, the search result data generation unit 124 determines drawing c2 as candidate 2. In this manner, the search result data generation unit 124 selects drawing c2 as the past drawing with the second highest relevance to drawing ND.

[0091] Furthermore, SCORE(a1) is the smallest relevance score (the Mth largest relevance score) among SCORE(a1) to SCORE(M1). Therefore, the search result data generation unit 124 determines drawing a1 as candidate M. In this manner, the search result data generation unit 124 selects drawing a1 as the past drawing having the lowest relevance (the Mth highest) to drawing ND.

[0092] The search result data generating unit 124 generates data indicating the candidate numbers of the drawings a1 to M1 for the drawing ND. In this data, the drawing numbers (identifiers) of the drawings a1 to M1 are sorted according to the SCORE. More specifically, in this data, the drawing numbers of the drawings a1 to M1 are sorted in descending order of SCORE.

[0093] Note that even if the type of the new drawing is different from the outline drawing, the search result data generating unit 124 performs the same process for the past drawings of the same type as the new drawing. For example, consider a case where the type of the new drawing is an assembly drawing. In this case, the search result data generating unit 124 determines the candidate numbers of the drawings a2 to M2 (assembly drawings) for the new drawing based on SCORE(a2) to SCORE(M2). The same applies to other types of drawings.

[0094] As described above, the search result data generation unit 124 sets candidate numbers for drawings (i,1) to (i,M) for drawings ND belonging to type i based on SCORE(i,1) to SCORE(i,N). Specifically, the search result data generation unit 124 identifies SCORE(i,g), which is the gth largest SCORE, from SCORE(i,1) to SCORE(i,M). Then, the search result data generation unit 124 determines the past drawing corresponding to SCORE(i,g), i.e., drawing (i,g), as candidate g.

[0095] As an example, the search result data generating unit 124 generates a fifth table TB5 as the search result data. In TB5, the past drawings are sorted in descending order of candidate number (in descending order of SCORE) for each type of drawing ND. An example of TB5 is shown in Fig. 8. The search result data generating unit 124 supplies TB5 to the display unit 72 and causes the display unit 72 to display the TB5.

[0096] By displaying TB5 on the display unit 72, the search results of the drawing search device 12 (more specifically, the search result data generating unit 124) can be presented to the user in a highly viewable form.

[0097] However, the search result data generating unit 124 does not necessarily have to present the entire TB5 to the user as the search result data. This is because, when the number of past drawings is too large, presenting the entire TB5 to the user may rather reduce the user's convenience.

[0098] Therefore, the search result data may be at least a part of TB5. The search result data may be data indicating at least one drawing (at least one past drawing) corresponding to the target drawing (drawing ND).

[0099] As described above, the drawing of candidate 1 is the drawing with the highest SCORE among the multiple past drawings. Therefore, from the viewpoint of reducing the amount of data of the search result data, it is possible to generate the search result data as data showing only the drawing of candidate 1.

[0100] However, when designing equipment in the field of plant engineering, designers need to make a comprehensive judgment based on various technical information. For this reason, when designing a new piece of equipment that is listed on Drawing ND using past drawings as reference, the past drawing with the highest SCORE (Candidate 1 drawing) is not necessarily the most suitable design reference. For example, Candidate 2 drawing (drawing with a lower SCORE) may actually be more suitable as a design reference than Candidate 1 drawing.

[0101] Therefore, it is preferable that the search result data is data showing multiple top candidates (for example, up to the γth candidate). That is, the search result data may be data showing drawings of candidates 1 to γ, where γ is an integer greater than or equal to 1 and less than or equal to g.

[0102] (Effects of the reference form) In plant engineering, as a project for object T progresses, it may be necessary to search for past drawings that are highly related to drawing ND (target drawing) (see also Patent Document 1). This is because in the above project, various plans (e.g., forecast of production period, trial calculation of production cost) for equipment of the same type in object T are often made with reference to the design and production track record of a specific type of equipment in past objects.

[0103] Here, if the past drawing contains similar specification information as the drawing ND, it is expected that the drawing will be highly relevant to the drawing ND. However, in plant engineering, it is common for the number of drawings of past projects to be huge. For this reason, even if the drawing search system of Patent Document 1 is used, it is not necessarily easy for a user to quickly find a desired drawing (a past drawing that is highly relevant to the drawing ND) from the search results. This is because, when a keyword-less search is performed in the drawing search system of Patent Document 1, a considerable number of past drawings are likely to be hit as search results.

[0104] Therefore, one idea is to use a known drawing search system having a keyword search function to search for a desired drawing that is highly related to the drawing ND. For example, it is possible to search for past drawings by using a specific character string described in the drawing ND as a keyword (search word). However, even when such a keyword search is performed, it is not necessarily easy for a user to quickly find the desired drawing from the search results. This is because even when a keyword search is performed, a considerable number of past drawings may be hit as search results. In addition, if a user cannot determine an appropriate search word, it is difficult to use a keyword search in the first place.

[0105] As described above, it is difficult for a user who does not have sufficient knowledge of plant engineering (e.g., a new designer) to quickly search for a desired drawing using a conventional drawing search system. Unless the user is well versed in plant engineering (e.g., a veteran designer), it is not necessarily easy to efficiently search for a desired drawing using a conventional drawing search system.

[0106] In view of such problems of the conventional technology, the inventors of the present application (hereinafter, the inventors) have newly created an information processing system 100s (more specifically, an information processing device 1s). The information processing device 1s can search for a desired drawing corresponding to a drawing ND by using a learning model obtained by machine learning for each past drawing.

[0107] First, in the learning device 11s of the information processing device 1s, a content parameter (search target drawing content parameter) related to a specific character string is set for each past drawing. Then, a weighting coefficient for each past drawing is set based on the search target drawing content parameter.

[0108] Then, in the drawing search device 12 of the information processing device 1s, a relevance score, which is an index showing the degree of relevance between the drawing ND and each past drawing, is calculated for each past drawing using the content parameters (target drawing content parameters) of the drawing ND and the weighting coefficient of each past drawing. Then, each past drawing is sorted in descending order of relevance score. That is, at least one past drawing as a candidate for the desired drawing is presented to the user based on the relevance score.

[0109] Thus, the drawing search device 12 allows the user to efficiently search for a desired drawing by keyword-less search, unlike known drawing search systems with keyword search functions. In particular, the drawing search device 12 allows the user to perform an efficient drawing search even when the user cannot determine an appropriate search word. For this reason, the drawing search device 12 is particularly suitable for improving the work efficiency of, for example, the above-mentioned new designers.

[0110] The drawing search system of Patent Document 1 is common to the information processing device 1s (more specifically, the drawing search device 12) in that it is based on keyword-less search. However, unlike the information processing device 1s, the drawing search system of Patent Document 1 does not order the search results. Therefore, in the drawing search system of Patent Document 1, past drawings that are ranked high in the search results do not necessarily have a high relevance to the drawing ND.

[0111] In contrast, in the information processing device 1s, data in which each past drawing is sorted in descending order of relevance score is presented to the user as a search result. Therefore, the user can confirm each past drawing presented as a search result in descending order of relevance score. In other words, the user can preferentially confirm the past drawings that are ranked as higher candidates.

[0112] In this way, unlike the drawing search system of Patent Document 1, the information processing device 1s does not require the user to check each past drawing presented as a search result one by one (using a brute force method). Therefore, even compared with the drawing search system of Patent Document 1, the information processing device 1s allows the user to find the desired drawing corresponding to the drawing ND more easily than before. That is, compared with the drawing search system of Patent Document 1, the information processing device 1s can further improve the user's convenience.

[0113] Incidentally, in recent years, image matching technology (e.g., face recognition technology) using machine learning has been widely used. It is also conceivable to use this image matching technology to search for a desired drawing. However, in the above image matching technology, the relevance between the drawing ND and each past drawing is determined solely based on the "appearance" (e.g., layout) as an image. Therefore, a drawing search method based on image matching technology (hereinafter, image matching-based drawing search) may not be suitable for drawing search compared to the information processing device 1s. This point will be described below.

[0114] For example, if the format of drawings is not standardized (unified), the way drawings are written may differ depending on the creator of each drawing. Therefore, even if the same content (e.g., specification information) is described in two drawings, the content may be written in different positions in each drawing. For example, see "Area L1 in FIG. 3" (hereinafter also referred to as Area A) and "Area R1 in FIG. 6" (hereinafter also referred to as Area B).

[0115] In such a case, in image matching-based drawing search, drawing a1 may be evaluated as having low relevance to drawing ND because the positions of area A and area B are different. However, as described above, the same type of specification information (specification information related to voltage values) is described in area A and area B. Therefore, in reality, drawing a1 may be considered to have a certain degree of relevance to drawing ND.

[0116] As described above, in the image matching-based drawing search, it is not possible to evaluate the relevance between drawing ND and drawing a1 by considering the content of the specification information described in each drawing. Therefore, it is not possible to properly search for a past drawing (drawing b1 in the example of Figure 8) that looks different from drawing ND but contains similar content.

[0117] In contrast, unlike the image matching-based drawing search, the information processing device 1s focuses on a specific character string described in each drawing and sets content parameters (search target drawing content parameters and target drawing content parameters). That is, the information processing device 1s can take into consideration the content itself of the specification information described in each drawing. Then, based on the content parameters, the relevance between the drawing ND and each past drawing is evaluated (e.g., scored).

[0118] Therefore, according to the information processing device 1s, unlike the image matching-based drawing search, it is possible to appropriately search for a past drawing (e.g., drawing b1) that has a different appearance from the drawing ND but has similar contents. In this way, the information processing device 1s is more suitable for drawing search than the image matching-based drawing search.

[0119] In addition, the more the number of past drawings, the more difficult it becomes to properly search for past drawings that are highly related to a given new drawing using a conventional drawing search system (or image matching based drawing search). Therefore, the more the number of past drawings, the more useful the information processing device 1s is for improving user convenience.

[0120] [Modifications] (1) The information processing device according to one aspect of the present invention can also be applied to searching for drawings in fields other than plant engineering. The drawings according to one aspect of the present invention may be drawings from which content parameters can be obtained by the drawing searching device.

[0121] (2) In the reference embodiment, the learning model generating unit 113s generates the learning model using multinomial logistic regression analysis. However, the method of generating the learning model is not limited to this.

[0122] The learning model generation unit 113s may generate a learning model by machine learning using the above-mentioned teacher data (may calculate each weighting coefficient). For example, the learning model generation unit 113s may generate a learning model by using a known neural network technique (e.g., a known deep learning technique).

[0123] (3) The search result data generating unit 124 may further acquire predetermined data corresponding to the drawing selected as the candidate g. In the example of Fig. 8, the search result data generating unit 124 may acquire specification data of a predetermined device corresponding to the drawing b1 (candidate 1). This is because in various plans of plant engineering, the specifications of a predetermined device described in the past drawings are often referred to together with the past drawings.

[0124] In this case, the search result data generating unit 124 can present the specification data to the user together with the search result data. This can further improve user convenience. As an example, a table showing the correspondence between the drawings a1 to MN and the specification data may be prepared in advance in the information processing device 1s. The search result data generating unit 124 may refer to the table and acquire the specification data corresponding to the candidate g.

[0125] [Embodiment 1] 9 is a block diagram showing the configuration of a main part of the information processing system 100 of the first embodiment. An information processing device of the information processing system 100 is referred to as an information processing device 1. A control device of the information processing device 1 is referred to as a control device 10. A learning device of the control device 10 is referred to as a learning device 11. First, prior to a specific description of the processing of the information processing system 100, possible improvements of the information processing system 100s of the reference embodiment will be described.

[0126] As described above, in the information processing device 1s of the reference embodiment, all past drawings (drawings a1 to MN) are used for learning. Therefore, in the information processing device 1s, it is considered that the quality of the learning model varies depending on the contents of the past drawings (drawings a1 to MN) stored in the past property drawing DB 91. As described below, if the contents of the past drawings are biased, a learning model based on the biased drawing contents (biased learning model) is generated.

[0127] (Pattern classification regarding the quality of learning models in reference form) The inventors have examined the quality of the learning model in the reference embodiment. Specifically, the inventors have classified the relationship between the past drawings and the quality of the learning model into the following four patterns.

[0128] In the following explanation of past drawings (e.g., production drawings), "contents linked to the drawing number" typically means "contents of specification drawings (drawings with specifications) for creating past drawings." Examples of production drawings include outline drawings and assembly drawings. However, as mentioned above, please note that past drawings are not limited to production drawings.

[0129] In the following example, for simplicity, it is assumed that the specification drawings for one item contain all the specifications for creating all the production drawings for that item. And, it is assumed that all the production drawings for the above item are created based on the specification drawings for that item.

[0130] In the following example, if a production drawing belonging to a certain drawing type (e.g. outline drawing) and another production drawing have the same (common) specifications for the certain drawing type (e.g. specifications for creating the outline drawing), the same drawing number will be assigned to the production drawing and the other production drawing. However, naturally, if the specifications for the certain drawing type are different between the production drawing and the other production drawing, different drawing numbers will be assigned to the production drawing and the other production drawing.

[0131] (Pattern 1: When the past drawings are composed only of past drawings with different contents linked to the drawing numbers and different drawing numbers) Different drawing contents can be learned from multiple past drawings with different drawing numbers. This allows the generation of particularly high-quality learning models (learning models with sufficient generalization). As a result, high search accuracy is achieved.

[0132] (Pattern 2: When the past drawings contain different contents linked to the drawing numbers, but the same drawing numbers are mixed in.) Different drawing contents are learned from drawings with the same drawing number. That is, learning is performed using ambiguous data. However, in general, machine learning can generate a high-quality learning model even when learning using such ambiguous data. Therefore, in pattern 2, as in pattern 1, high search accuracy is achieved.

[0133] (Pattern 3: When the past drawings contain drawings with the same or similar content linked to the drawing numbers but different drawing numbers) The same or similar drawing contents are learned from multiple past drawings with different drawing numbers. That is, biased learning is performed. For this reason, in pattern 3, a biased learning model is generated. As a result, search accuracy may decrease. For example, there is a high possibility that multiple different drawings having the same or similar drawing contents will be hit as candidates 1 to g.

[0134] (Pattern 4: When the past drawings contain drawings with the same or similar content linked to the drawing number and the same drawing number) The same or similar drawing contents are learned multiple times using past drawings with the same drawing number. In this case, biased learning is also performed. For this reason, a biased learning model is generated even in pattern 4. As a result, the search accuracy may decrease.

[0135] As can be seen from patterns 3 and 4 above, a biased learning model can lead to a decrease in search accuracy. For this reason, a biased learning model can also be considered a low-quality learning model. If a drawing search is performed using such a biased learning model, the possibility of obtaining appropriate search results may decrease.

[0136] Therefore, the inventors created the information processing system 100 of the first embodiment in order to further improve the information processing system 100s. As described below, the information processing system 100 (more specifically, the learning device 11) can prevent the quality of the learning model from deteriorating. Therefore, it is possible to achieve a higher search accuracy than the information processing system 100s.

[0137] (Learning Device 11) Unlike the learning device 11s, the learning device 11 further includes a dissimilarity calculation unit 114, a similar set determination unit 115, and a non-learning target drawing determination unit 116. The similar set determination unit 115 and the non-learning target drawing determination unit 116 may be collectively referred to as a determination unit. Moreover, the learning device 11 includes a learning model generation unit 113 (learning unit) instead of the learning model generation unit 113s of the learning device 11s.

[0138] (Dissimilarity calculation unit 114) The dissimilarity calculation unit 114 calculates the dissimilarity between any two past drawings based on the content parameters (first to Lth content parameters) of the two past drawings derived by the past drawing data preprocessing unit 112.

[0139] In the following description, any two past drawings will be referred to as drawing I and drawing J. I and J are integers greater than or equal to 1 and less than or equal to T. For convenience, in the following description, the k-th content parameter of drawing I will be referred to as Ik. Similarly, the k-th content parameter of drawing J will be referred to as Jk. Furthermore, the degree of difference between drawing I and drawing J will be referred to as D(I,J). D(I,J) is an index value indicating the degree of difference between the contents (description) of drawing I and drawing J.

[0140] FIG. 10 is a diagram for explaining the degree of difference in the first embodiment. In FIG. 10, for example, "drawing a1" is abbreviated as "a1". Similar abbreviations are used appropriately in this specification. FIG. 10 shows an L-dimensional data space (L-dimensional space) showing the distribution of content parameters of each drawing. In the L-dimensional space of FIG. 10, point I corresponding to drawing I (point indicating the content parameters of drawing I) is called P(I). The coordinates of P(I) are P(I) = [I1,I2,…,Ik,…,IL] The dissimilarity calculation unit 114 acquires I1 to IL from the previous drawing data preprocessing unit 112, and thereby determines the coordinates of P(I) in the L-dimensional space.

[0141] Similarly, in the L-dimensional space of FIG. 10, point J corresponding to drawing J is called P(J). The coordinates of P(J) are as follows: P(J) = [J1, J2, …, Jk, …, JL] The dissimilarity calculation unit 114 acquires J1 to JL from the past drawing data preprocessing unit 112, and thereby determines the coordinates of P(J) in the L-dimensional space.

[0142] As an example, the dissimilarity calculation unit 114 calculates the distance between point I and point J in the L-dimensional space as D(I,J). In the first embodiment, the dissimilarity calculation unit 114 calculates the Euclidean distance between point I and point J as D(I,J). That is, the dissimilarity calculation unit 114 calculates D(I,J)={(I1-J1) 2 +…+(IL-JL) 2} 1 / 2 = {Σ(Ik-Jk) 2} 1 / 2 …(4) Calculate D(I,J) as follows.

[0143] As is clear from formula (4) above, D(I,J) takes a value greater than or equal to 0. The larger D(I,J) is, the lower the similarity in content between drawing I and drawing J. Conversely, the smaller D(I,J) is, the higher the similarity in content between drawing I and drawing J. When D(I,J)=0, the similarity between drawing I and drawing J is at its maximum. Specifically, when D(I,J)=0, drawing I and drawing J are drawings with the same content.

[0144] The dissimilarity calculation unit 114 calculates D(I, J) for all combinations of I and J. Then, the dissimilarity calculation unit 114 generates a table indicating the dissimilarity between each of the drawings a1 to MN as a dissimilarity table TB11. TB11 includes an i-th sub-table TB11-i in the dissimilarity table. TB11-i is a table indicating the dissimilarity between each past drawing belonging to the drawing i. FIG. 11 shows TB11-1 as an example. TB11-1 shows the dissimilarity between each past drawing (drawings a1 to M1) belonging to the drawing 1 (outline drawing). TB11-1 is a square matrix with M rows and M columns. TB11 is composed of a set of N sub-tables, TB11-1 to TB11-N. In this way, in the first embodiment, a sub-table for the dissimilarity is created for each drawing type.

[0145] Hereinafter, with reference to TB11-1, a description will be given of the operations of the dissimilarity calculation unit 114, the similar set determination unit 115, and the non-learning drawing determination unit 116. However, the following description of each item related to TB11-1 also applies to TB11-2 to TB11-N in the same manner.

[0146] As is clear from the above formula (4), TB11-1(I,J) = TB11-1(J,I). Also, TB11-1(I,I) = 0. Therefore, as shown in FIG. 11, TB11-1 is generated as a symmetric matrix with diagonal components of 0. Thus, in the following description, the case where I < J will be mainly described.

[0147] The dissimilarity calculation unit 114 sets the value of TB11-1(I,J) such that TB11-1(I,J) = D(I,J). As an example, assume that D(a1,b1) = 8.1. In this case, the dissimilarity calculation unit 114 sets the value of TB11-1(a1,b1) = 8.1 and sets the value of TB11-1(1,2).

[0148] (Similar set determination unit 115) The similar set determination unit 115 determines a set of past drawings (similar set) that are considered to have high similarity among the drawings a1 to M1 based on each dissimilarity shown in TB11-1. Specifically, the similar set determination unit 115 determines the similar set based on a dissimilarity threshold value (for convenience, referred to as Dth). In Embodiment 1, the case where Dth = 6.0 is set as an example. However, Dth may be arbitrarily set by the user of the information processing system 100.

[0149] FIG. 12 is a diagram showing an example of the process of the similar set determination unit 115. In the example of FIG. 12, for convenience of explanation, TB11-1 in FIG. 11 is also referred to as table 1200A. Table 1200A may be referred to as the first-generation dissimilarity table.

[0150] The similar set determination unit 115 searches for a component having the minimum dissimilarity from the table 1200A under the condition of I < J. Subsequently, the similar set determination unit 115 determines whether the discovered minimum dissimilarity (referred to as Dmin for convenience) is less than or equal to Dth. If Dmin is less than or equal to Dth, the similar set determination unit 115 determines the set of past drawings (set of past drawings) corresponding to Dmin as a similar set. In other words, when Dmin is less than or equal to Dth, the similar set determination unit 115 classifies (groups) the drawings a1 to M1 so that the set of past drawings corresponding to Dmin belongs to the same set.

[0151] In the example of FIG. 12, as a result of the search for the table 1200A, the similar set determination unit 115 specifies D(c1, e1) = 2.5 as the minimum component (see the hatched part of the table 1200A). Therefore, the similar set determination unit 115 sets Dmin = 2.5. In the table 1200A, Dmin < Dth. For this reason, the similar set determination unit 115 determines the set {c1, e1} as the set A (see also FIG. 13 described later). The set A may be referred to as the first similar set.

[0152] Subsequently, the similar set determination unit 115 generates a table 1200B by integrating the elements c1 and e1 (components of the set A) in the table 1200A into the set A. The table 1200B may be referred to as the second-generation dissimilarity table. When generating the table 1200B, the similar set determination unit 115 sets the dissimilarities between the set A and each element outside the set A (each past drawing), that is, D(A, a1), D(A, b1),..., D(A, M1), by a predetermined method.

[0153] In Embodiment 1, the similar set determination unit 115 sets the maximum value of the dissimilarities between each element belonging to the set A (that is, the drawings c1 and e1) and one element not belonging to the set A (referred to as the drawing X for convenience) as D(A, X). That is, the similar set determination unit 115 D(A, X) = Max{D(c1, X), D(e1, X)} …(5) As D(A, X) is set. Thus, in Embodiment 1, D(A, X) is set by the farthest distance method of the hierarchical clustering. The same applies to the set B described later.

[0154] As an example, consider the case where X = a1. As shown in Table 1200B of FIG. 12, the similar set determination unit 115 uses the above formula (5) to D(A, a1) = Max{D(c1, a1), D(e1, a1)} = Max(3.3, 10.3) = 10.3 and set D(A, a1). Similarly, the similar set determination unit 115 sets D(A, b1), …, D(A, M1).

[0155] Subsequently, the similar set determination unit 115 searches the table 1200B for the component with the minimum dissimilarity under the condition of I < J. As a result of the search for the table 1200B, the similar set determination unit 115 specifies D(b1, d1) = 5.8 as the minimum component (see the hatched part of the table 1200B). Therefore, the similar set determination unit 115 sets Dmin = 5.8 as Dmin. In the table 1200B, Dmin < Dth. For this reason, the similar set determination unit 115 determines the set {b1, d1} as the set B (see also FIG. 13). The set B may be referred to as the second similar set.

[0156] Subsequently, the similar set determination unit 115 generates a table 1200C by integrating the elements b1 and d1 (components of the set B) in the table 1200B into the set B. The table 1200C may be referred to as the third-generation dissimilarity table. The similar set determination unit 115 uses the above formula (5) to set the dissimilarity between the set B and each element outside the set B, that is, D(B, A), D(B, a1), …, D(B, M1). However, it should be noted that in the calculation of (B, A), D(B, a1), …, D(B, M1), the set A on the left side of the formula (5) is replaced by the set B.

[0157] As an example, the similar set determination unit 115 D(B,A)=Max{D(b1,A),L(d1,A)} =Max(12.8,11.4) =12.9 sets D(B,A) as such. As can be understood from the calculation example of D(B,A), in the calculation by the similar set determination unit 115, one set (e.g., set A) is regarded as one virtual past drawing representing a plurality of past drawings (drawings c1 and e1).

[0158] Subsequently, the similar set determination unit 115 searches the table 1200C for the component having the minimum dissimilarity under the condition of I < J. As a result of the search for the table 1200C, the similar set determination unit 115 specifies D(B,a1)=8.1 as the minimum component (see the hatched part of the table 1200C). Therefore, the similar set determination unit 115 sets Dmin to 8.1. In the table 1200C, Dmin > Dth. When Dmin is greater than Dth, the similar set determination unit 115 completes the set grouping process.

[0159] FIG. 13 is an example of a dendrogram DG showing the hierarchical structure of the sets obtained in Embodiment 1. By using the DG, the user can visually recognize the hierarchical structure of the sets. The DG may be used as auxiliary information for the user to adjust Dth.

[0160] As also understood from FIG. 13, when Dth is set to 8.1 or more, the similar set determination unit 115 classifies the drawing a1 and the set B into the same set (for convenience, referred to as set B'). As described above, B = {b1, d1}. Therefore, B' = {a1, b1, d1} = {a1, {b1, d1}} = {a1, B} can also be expressed as such. Thus, in Embodiment 1, it is also possible to handle a new set (e.g., set B') in which an existing set (e.g., set B) is nested.

[0161] Here, referring back to FIG. 10, a case where D(B', e1) is calculated will be considered. In the first embodiment, the similar set determination unit 115 calculates D(B', e1) using the following formula (6): D(B', e1) =Max{D(a1,e1),D(b1,e1),D(d1,e1)}…(6) That is, the similar set determination unit 115 may calculate L(B', e1) by calculating the maximum distance between each drawing belonging to the set B' and the drawing e1. may be determined.

[0162] However, the above formula (6) is D(B', e1) =Max{D(a1,e1),Max{D(b1,e1),D(d1,e1)}} =Max{D(a1,e1),D(B,e1)}…(6') Therefore, the similar set determination unit 115 can calculate D(B', e1) by using equation (6') instead of equation (6).

[0163] In formula (6'), the number of arguments of the Max function is smaller than that in formula (6). Therefore, by using formula (6') instead of formula (6), the calculation of D(B', e1) in the similar set determination unit 115 can be accelerated. That is, by using formula (6'), D(B', e1) can be calculated in a shorter time than when formula (6) is used. In this way, in the first embodiment, various calculations can be performed faster (more efficiently) by handling a nested set structure.

[0164] (Non-learning drawing determination unit 116) The non-learning drawing determination unit 116 determines non-learning drawings (past drawings not used for learning in the learning model generation unit 113) from among the drawings a1 to M1 according to the degree of difference between the drawings a1 to M1. In the first embodiment, the non-learning drawing determination unit 116 determines non-learning drawings from among the past drawings belonging to each set (e.g., set A and set B) determined by the similar set determination unit 115.

[0165] 14 is a diagram showing an example of the processing of the non-learning drawing determination unit 116. In the first embodiment, the non-learning drawing determination unit 116 determines one representative drawing (representative drawing in a set) from among a plurality of drawings belonging to a certain set (for convenience, referred to as set X). Then, the non-learning drawing determination unit 116 determines each drawing, excluding the representative drawing, from among the plurality of drawings belonging to set X as a non-learning drawing. The determination of the representative drawing may be performed by any method. As an example, the non-learning drawing determination unit 116 may determine the drawing with the latest timestamp from among the plurality of drawings belonging to set X as the representative drawing.

[0166] FIG. 14 illustrates an example in which the non-learning drawing determination unit 116 (i) determines drawing c1 as a representative drawing in set A from among two drawings c1 and e1 belonging to set A, and (ii) determines drawing b1 as a representative drawing in set B from among two drawings b1 and d1 belonging to set B. In this case, the non-learning drawing determination unit 116 (i) determines drawing e1 as a non-learning drawing in set A, and (ii) determines drawing d1 as a non-learning drawing in set B. In this way, in the example of FIG. 14, drawings d1 and e1 are determined as non-learning drawings from among drawings a1 to M1.

[0167] After completing the above-mentioned processing for drawing 1 (drawings a1 to M1), the dissimilarity calculation unit 114, similar set determination unit 115, and non-learning drawing determination unit 116 perform the same processing for drawing 2 (drawings a2 to M2). That is, the dissimilarity calculation unit 114, similar set determination unit 115, and non-learning drawing determination unit 116 perform the above-mentioned processing in the order of drawing 1 → drawing 2 → ... → ... drawing N. In this way, the non-learning drawing determination unit 116 determines non-learning drawings from among the drawings a1 to MN according to the dissimilarity between the drawings a1 to MN.

[0168] Then, the non-learning drawing determination unit 116 edits the above-mentioned TB2 according to the non-learning drawing determined by itself. Specifically, the non-learning drawing determination unit 116 edits TB2-i according to the non-learning drawing determined in drawing i. FIG. 14 shows an example of editing TB2-1 by the non-learning drawing determination unit 116. In the example of FIG. 14, the non-learning drawing determination unit 116 deletes content parameters corresponding to the non-learning drawings in drawing 1 (i.e., drawings d1 and e1) from TB2-1. In addition, the non-learning drawing determination unit 116 assigns the drawing number of drawing c1 (representative drawing of set A) to set A, and assigns the drawing number of drawing b1 (representative drawing of set B) to set B.

[0169] The non-learning drawing determination unit 116 generates the edited second table TB12 by editing TB2 as described above. Specifically, the non-learning drawing determination unit 116 generates the i-th sub-table TB12-i in the edited second table by editing TB2-i as described above. TB12-1 is illustrated in FIG. 15. TB12 is composed of a set of N sub-tables, TB12-1 to TB12-N.

[0170] (Learning model generation unit 113) The learning model generation unit 113 acquires the TB12 generated by the non-learning drawing determination unit 116 as teacher data. Then, the learning model generation unit 113 generates a learning model by performing a learning process (e.g., multinomial logistic regression analysis) similar to that of the reference embodiment. Each process after the generation of the learning model is similar to that of the reference embodiment.

[0171] (Effects of the First Embodiment) Similarly to the reference embodiment, the information processing system 100 (information processing device 1) can improve the user's convenience in searching for drawings more than ever before. In addition, unlike the reference embodiment, the learning device 11 can generate a learning model using the TB 12. That is, unlike the reference embodiment, the learning device 11 can generate a learning model based on the content parameters of each past drawing excluding drawings that are not the learning target.

[0172] As described above, in the learning device 11, drawings that are highly similar to other past drawings are determined as drawings not to be learned. Then, the drawings not to be learned can be excluded and learning can be performed based on the past drawings. In other words, learning can be performed based on a group of past drawings whose contents are low in similarity to each other. Therefore, the possibility of learning the above-mentioned patterns 3 and 4 can be reduced. As a result, the possibility of obtaining a biased learning model can be reduced.

[0173] In this way, the learning device 11 can prevent the quality of the learning model from deteriorating by performing learning while excluding drawings that are not the learning target (ensuring the quality of the learning model). Therefore, in the information processing system 100, a search can be performed by the drawing search device 12 using a high-quality learning model generated by the learning device 11. As a result, the information processing system 100 can achieve even higher search accuracy than the reference embodiment.

[0174] [Modifications] (1) In the first embodiment, the distance in the L-dimensional space is exemplified as an example of the dissimilarity according to one aspect of the present invention. In particular, in the first embodiment, the Euclidean distance is exemplified as an example of the distance. However, the distance is not limited to the Euclidean distance. Other examples of the distance include the Manhattan distance and the Chebyshev distance. The distance may be any known type of distance (norm) that can be defined in the L-dimensional space.

[0175] In addition, it should be noted that the dissimilarity according to one embodiment of the present invention is not necessarily limited to the distance. D(I,J) may be any index value calculated based on I1-IL and J1-JL, as long as it can express the difference between the contents of drawings I and J.

[0176] (2) In the first embodiment, the case where D(A,X) is set by the longest distance method is illustrated. However, the method of setting D(A,X) is not limited to the longest distance method. For example, D(A,X) may be set by other methods such as the shortest distance method, the center of gravity method, or the group average method.

[0177] When the longest distance method is used, D(A,X) can be set to a larger value than when the other methods described above are used. Therefore, by using the longest distance method, it is expected that the classification process of similar sets can be completed with a smaller number of calculations in the similar set determination unit 115. Therefore, the longest distance method can be said to be a suitable method for reducing the calculation cost of the learning device 11.

[0178] [Software implementation example] The control block (particularly the control device 10) of the information processing system 100 may be realized by a logic circuit (hardware) formed on an integrated circuit (IC chip) or the like, or may be realized by software.

[0179] In the latter case, the information processing system 100 includes a computer that executes instructions of a program, which is software that realizes each function. The computer includes, for example, one or more processors, and a computer-readable recording medium that stores the program. The object of the present invention is achieved by the processor reading the program from the recording medium and executing it in the computer. The processor may be, for example, a CPU (Central Processing Unit). The recording medium may be a "non-transient tangible medium," such as a ROM (Read Only Memory), tape, disk, card, semiconductor memory, programmable logic circuit, or the like. The information processing system may further include a RAM (Random Access Memory) that expands the program. The program may be supplied to the computer via any transmission medium (such as a communication network or broadcast waves) that can transmit the program. Note that one aspect of the present invention may also be realized in the form of a data signal embedded in a carrier wave, in which the program is embodied by electronic transmission.

[0180] [Additional Notes] The present invention is not limited to the above-described embodiments, and various modifications are possible within the scope of the claims. Embodiments obtained by appropriately combining the technical means disclosed in different embodiments are also included in the technical scope of the present invention. [Explanation of symbols]

[0181] 1. Information processing device 10 Control device 11 Learning device (model generation device) 100 Information Processing Systems 111 Past Drawing Data Acquisition Department 112 Past drawing data preprocessing section (search target drawing content parameter acquisition section) 113 Learning model generation unit (learning unit) 114 Dissimilarity calculation unit 115 Similar set determination unit (determination unit) 116 Drawings not to be studied decision unit (decision unit) a1~MN Drawings (past drawings, drawings to be searched) ND drawing (new drawing, target drawing) A,B set d1, e1 Drawings (drawings not included in the study)

Claims

1. A model generation device that generates a learning model for searching at least one drawing corresponding to a target drawing from a plurality of search target drawings, a search target drawing content parameter acquisition unit that acquires search target drawing content parameters related to the description contents of each of the plurality of search target drawings by analyzing the plurality of search target drawings; a dissimilarity calculation unit that calculates a dissimilarity between any two different search target drawings included in the plurality of search target drawings based on the search target drawing content parameters of each of the plurality of search target drawings; a determining unit that determines non-learning drawings from among the plurality of search target drawings according to the degree of difference; A learning unit that generates the learning model based on the search target drawing content parameters of each of the plurality of search target drawings excluding the non-learning target drawings, The determination unit is determining a set consisting of pairs of the plurality of search target drawings whose dissimilarity is equal to or less than a threshold value; determining the non-learning drawings from among the plurality of search target drawings belonging to the set; The determination unit sets the maximum value of the dissimilarity between (i) each of the plurality of search target drawings belonging to the set and (ii) a certain search target drawing that does not belong to the set as the dissimilarity between the set and the certain search target drawing, a model generating device.

2. The search target drawing content parameters include L parameters from a search target drawing first content parameter to a search target drawing Lth content parameter, One of the two different search target drawings is called drawing I, and the other is called drawing J. The model generating device according to claim 1, wherein the dissimilarity calculation unit calculates the distance between a point corresponding to drawing I and a point corresponding to drawing J in an L-dimensional data space indicating the distribution of the search target drawing content parameters as the dissimilarity between drawing I and drawing J.

3. A model generation method for generating a learning model for searching at least one drawing corresponding to a target drawing from a plurality of search target drawings, comprising: a search target drawing content parameter acquisition step of acquiring search target drawing content parameters related to the contents of each of the plurality of search target drawings by analyzing the plurality of search target drawings; a difference calculation step of calculating a difference in contents between any two different search target drawings included in the plurality of search target drawings based on the search target drawing content parameters of each of the plurality of search target drawings; a determining step of determining non-learning drawings from among the plurality of search target drawings according to the degree of difference; A learning process for generating the learning model based on the search target drawing content parameters of each of the plurality of search target drawings excluding the non-learning target drawings, In the above determination step, determining a set consisting of pairs of the plurality of search target drawings whose dissimilarity is equal to or less than a threshold value; determining the non-learning drawings from among the plurality of search target drawings belonging to the set; A model generation method in which, in the determination process, the maximum value of the dissimilarity between (i) each of the plurality of search target drawings belonging to the set and (ii) one search target drawing that does not belong to the set is set as the dissimilarity between the set and the one search target drawing.

Citation Information

Patent Citations

  • Neural network learning method

    JP1992299442A

  • Drawing management system

    JP2003099484A

  • Abnormality detecting method and abnormality detecting system

    JP2010191556A

  • Drawing management system and drawing management method

    JP2020091769A