Analysis program, information processing device and analysis method

Abstract

To provide a novel technique using a large-scale language model for comparing input information representing a technical idea with other information.SOLUTION: An analysis program causes a computer to execute the steps of: acquiring input information representing a technical idea; acquiring comparative information of an object to be compared with the technical idea; dividing the technical idea for every component; generating a first instruction sentence including an instruction to compare each of the plurality of divided components with the comparative information; and outputting an evaluation result representing a relevance of each of the plurality of components to the comparative information, based on a result obtained from the large-scale language model by inputting the first instruction into the large-scale language model.SELECTED DRAWING: Figure 2

Description

Technical Field

[0001] The present disclosure relates to an analysis program, an information processing apparatus, and an analysis method.

Background Art

[0002] In recent years, various large language models have been developed. A large language model is a language model that has learned a vast amount of text data and has been trained to process various natural languages.

[0003] Non-Patent Document 1 discloses a method for improving the output accuracy of ChatGPT, which is an example of a large language model. Specifically, Non-Patent Document 1 discloses that by inputting an instruction sentence "Let's think step by step" into ChatGPT, the accuracy of the output is improved.

Prior Art Documents

Non-Patent Documents

[0004]

Non-Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0005] Although large language models can be applied to various services, there are still few examples of being realized as specific services. In this regard, a new technology that utilizes a large language model to compare input information indicating a technical idea with other information is desired.

Means for Solving the Problems

[0006] In an example of the present disclosure, an analysis program is provided. The analysis program causes a computer to perform steps of obtaining input information indicating a technical idea, obtaining comparison information to be compared with the technical idea, dividing the technical idea into constituent elements for each, generating a first instruction sentence including an instruction to compare each of the plurality of divided constituent elements with the comparison information, and outputting an evaluation result indicating the relevance between each of the plurality of constituent elements and the comparison information based on a result obtained from the large language model by inputting the first instruction sentence into the large language model.

[0007] In an example of the present disclosure, the dividing step includes generating a second instruction sentence for dividing the technical idea into constituent elements for each, and inputting the second instruction sentence into the large language model.

[0008] In an example of the present disclosure, the dividing step includes dividing the technical idea into constituent elements for each according to a predetermined rule.

[0009] In an example of the present disclosure, the first instruction sentence includes an instruction for causing the large language model to output the description positions of the plurality of constituent elements in the comparison information.

[0010] In an example of the present disclosure, the comparison information includes patent documents. The description location includes at least one of a paragraph number in the patent document, a figure number in the patent document, and a claim number in the patent document.

[0011] In an example of the present disclosure, the first instruction sentence includes an instruction for causing the large language model to output the reason for extracting the description position.

[0012] In an example of the present disclosure, the first instruction sentence includes an instruction for causing the large language model to output the degree of relevance between each of the plurality of constituent elements and the comparison information.

[0013] In one example of the present disclosure, the analysis program further causes the computer to generate a third instruction statement that summarizes the comparison information from a predetermined perspective, and output a summary regarding the comparison information based on a result obtained from the large language model by inputting the third instruction statement into the large language model.

[0014] In another example of the present disclosure, an information processing apparatus is provided. The information processing apparatus includes a control unit for controlling the information processing apparatus. The control unit executes a process of acquiring input information indicating a technical idea, a process of acquiring comparison information to be compared with the technical idea, a process of dividing the technical idea into constituent elements, a process of generating a first instruction statement including an instruction to compare each of the plurality of divided constituent elements with the comparison information, and a process of outputting an evaluation result indicating the relevance between each of the plurality of constituent elements and the comparison information based on a result obtained from the large language model by inputting the first instruction statement into the large language model.

[0015] In another example of the present disclosure, an analysis method executed by a computer is provided. The analysis method includes a step of acquiring input information indicating a technical idea, a step of acquiring comparison information to be compared with the technical idea, a step of dividing the technical idea into constituent elements, a step of generating a first instruction statement including an instruction to compare each of the plurality of divided constituent elements with the comparison information, and a step of outputting an evaluation result indicating the relevance between each of the plurality of constituent elements and the comparison information based on a result obtained from the large language model by inputting the first instruction statement into the large language model.

[0016] The above and other objects, features, aspects and advantages of the present invention will become apparent from the following detailed description of the present invention understood in connection with the accompanying drawings.

Brief Description of the Drawings

[0017]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Figure 11

Figure 12

Figure 13

Figure 14

Figure 15

Figure 16

Embodiments for Carrying Out the Invention

[0018] Hereinafter, each embodiment according to the present invention will be described with reference to the drawings. In the following description, the same parts and components are denoted by the same reference numerals. Their names and functions are also the same. Therefore, detailed descriptions thereof will not be repeated. Note that each embodiment and each modification described below may be selectively combined as appropriate.

[0019] <A. Information Processing System 10> First, with reference to FIG. 1, the device configuration of the information processing system 10 will be described. FIG. 1 is a diagram showing an example of the device configuration of the information processing system 10.

[0020] As shown in FIG. 1, the information processing system 10 includes an information processing device 100, a user terminal 200, and a server 300. The information processing device 100, the user terminal 200, and the server 300 are configured to be communicable with each other through a network NW (for example, the Internet).

[0021] The information processing device 100 is a notebook or desktop PC (Personal Computer), a tablet terminal, a smartphone, or other computer equipped with a communication function. The number of information processing devices 100 constituting the information processing system 10 may be one or two or more. The information processing device 100 is operated, for example, by company "A".

[0022] The user terminal 200 is, for example, a notebook or desktop PC, a tablet terminal, a smartphone, or other computer equipped with a communication function. The number of user terminals 200 constituting the information processing system 10 may be one or two or more. The information processing device 100 is owned, for example, by user "A" who is a general user.

[0023] The server 300 is a notebook or desktop PC (Personal Computer), a tablet terminal, a smartphone, or other computer with a communication function. The number of servers 300 constituting the information processing system 10 may be one or two or more. The server 300 is, for example, operated by company "B".

[0024] The server 300 stores a large language model 324. The large language model 324 is a language model that has learned an enormous amount of text data of several billion or more, and has been learned to be able to process various natural languages. The large language model 324 is also called an LLM (Large Language Models). The large language model 324 has been learned to generate an output corresponding to the instruction text when receiving an input of the instruction text.

[0025] Examples of the large language model 324 include the GPT series such as GPT-3 (Generative Pretrained Transformer) and GPT-4, PaLM (Scaling Language Modeling with Pathways), LLaMA (Large Language Model Meta AI), and known LLMs. In addition to the GPT series, various large language models such as Transformer-based large language models such as BERT (Bidirectional Encoder Representations from Transformers), CNN (Convolutional Neural Network), RNN (Recurrent Neural Network), and LSTM (Long Short Term Memory) may be used as the large language model.

[0026] Company "B", for example, has published an API (Application Programming Interface) for utilizing the functions of the large language model 324. As a result, designers and general users of Company "A" can utilize the functions of the large language model 324 through this API.

[0027] In addition, the various processes described in this specification may be implemented in the information processing apparatus 100, may be implemented in the user terminal 200, may be implemented in the server 300, or may be implemented in other computers.

[0028] Also, in the above description, an example in which the information processing system 10 includes the server 300 has been described, but the information processing system 10 may not include the server 300. In this case, the information processing system 10 is composed of one or more information processing apparatuses 100 and one or more user terminals 200.

[0029] <B. Outline of Processing> The information processing apparatus 100 provides the user "A" with a function for evaluating the relevance between input information indicating a technical idea and comparison information.

[0030] The "technical idea" is a technical means for solving a technical problem. The technical idea is defined by a combination of elements (hereinafter, also referred to as "constituent elements") that constitute an invention. Examples of the input information indicating the technical idea include, for example, the claims described in the claims of a patent, the constituent elements obtained by summarizing technical materials with a large language model, and the constituent elements obtained by summarizing an image with a large language model.

[0031] The "comparison information" is information to be compared with the above input information. The comparison information may be publicly known information in a state that can be publicly known to unspecified persons, or may be non-public information secretly managed within the company.

[0032] Examples of publicly available information include published patent documents and published non-patent documents. Examples of non-public information include, for example, information that is confidentially managed within a company (such as patent documents and technical materials before publication).

[0033] Examples of patent documents include, for example, published patent gazettes, patent gazettes, published patent applications, re-published patents, utility model gazettes, etc. As an example, patent documents include bibliographic data, specifications, claims, drawings, and abstracts. Examples of bibliographic data include, for example, application numbers, publication numbers, patent registration numbers, application dates, publication dates, registration dates, applicants, patent holders, invention titles, agents, and application countries. Examples of non-patent documents include, for example, academic papers, news articles, books, and web pages.

[0034] Hereinafter, "publicly available information" will be used as an example of "comparative information" for explanation, but comparative information is not limited to publicly available information.

[0035] With reference to Figure 2, the outline of the function for evaluating the relevance between input information 123 and publicly available information 125 will be described. Figure 2 is a diagram for explaining the relevance evaluation function.

[0036] The information processing apparatus 100 acquires input information 123 indicating a technical idea. The acquisition source of the input information 123 is arbitrary. As an example, the input information 123 is acquired from the above-described user terminal 200.

[0037] In addition, the information processing apparatus 100 acquires publicly available information 125 to be compared with the input information 123. The acquisition source of the publicly available information 125 is arbitrary. As an example, the publicly available information 125 is acquired from a patent database 124 (see Figure 3) stored in a storage device within the information processing apparatus 100, which will be described later. As another example, the publicly available information 125 is acquired from a patent database within a server managed by the Japan Patent Office. As yet another example, the publicly available information 125 is acquired from a patent database managed by another server.

[0038] The evaluation of the relevance between the input information 123 and the published information 125 is realized by using the large language model 324. More specifically, first, the information processing apparatus 100 divides the technical idea shown in the input information 123 into components one by one. Then, the information processing apparatus 100 generates an instruction sentence 126 (the first instruction sentence) including an instruction to compare each of the divided plurality of components with the published information 125.

[0039] The instruction sentence 126 is, for example, pre-registered in the information processing apparatus 100 as a template. The information processing apparatus 100 generates the instruction sentence 126 by specifying various information in the argument parts 127 to 129 in the instruction sentence 126.

[0040] More specifically, the acquired input information 123 is specified in the argument part 127. The obtained published information 125 is specified in the argument part 129.

[0041] Each of the divided components is sequentially specified in the argument part 128. For example, when the technical idea shown in the input information 123 is divided into N (N is an integer of 2 or more) components, the information processing apparatus 100 specifies each of the N components in the argument part 128. Thereby, the information processing apparatus 100 generates the instruction sentence 126 according to the number of the divided components. In the example of FIG. 2, N instruction sentences 126 are generated.

[0042] The generated instruction sentence 126 is input to the large language model 324. When receiving the input of the instruction sentence 126, the large language model 324 generates an answer according to the instruction sentence 126.

[0043] The information processing apparatus 100 outputs an evaluation result 130 indicating the relevance between each of the divided components and the published information 125 based on the result obtained from the large language model 324 by inputting the instruction sentence 126 to the large language model 324. The relevance may be represented by a numerical value indicating the degree of relevance or by an explanatory text. In the example of FIG. 2, the relevance is represented by a numerical value.

[0044] The output destination of the evaluation result 130 by the information processing apparatus 100 is arbitrary. As an example, the output destination is the user terminal 200. The evaluation result 130 output to the user terminal 200 is displayed on, for example, the display of the user terminal 200.

[0045] The above evaluation function is effective, for example, when conducting an invalid document investigation. More specifically, the user designates the patent invention to be invalidated as the input information 123. Thereafter, the information processing apparatus 100 decomposes the patent invention into constituent elements and compares each constituent element with the public information 125. Then, the information processing apparatus 100 outputs an evaluation result 130 indicating the degree to which each constituent element is disclosed in the public information 125. Thereby, the user can significantly shorten the time required for document screening and document review.

[0046] Note that the use of the above evaluation function is not limited to invalid document investigation. As another example, the above evaluation function may be used for prior art investigation. In this case, the user designates the invention to be the subject of the prior art investigation as the input information 123. Thereafter, the information processing apparatus 100 decomposes the invention into constituent elements and compares each constituent element with the public information 125. Then, the information processing apparatus 100 outputs an evaluation result 130 indicating the degree to which each constituent element is disclosed with the public information 125. Thereby, the user can easily determine whether the invention under investigation has novelty or inventiveness. As a result, the user can significantly shorten the time required for document screening and document review.

[0047] As described above, the information processing apparatus 100 utilizes the large language model 324 for the comparison between the input information 123 indicating the technical idea and the public information 125, and provides new value to the user.

[0048] <C. Hardware Configuration> Next, with reference to FIGS. 3 and 4, the hardware configurations of the information processing apparatus 100 and the user terminal 200 shown in FIG. 1 above will be described in order.

[0049] Note that the hardware configuration of the server 300 shown in FIG. 1 is the same as that of the information processing apparatus 100, and thus its description will not be repeated.

[0050] (C1. Information processing apparatus 100) First, with reference to FIG. 3, the hardware configuration of the information processing apparatus 100 shown in FIG. 1 will be described. FIG. 3 is a schematic diagram showing an example of the hardware configuration of the information processing apparatus 100.

[0051] The information processing apparatus 100 includes a control device 101, a ROM (Read Only Memory) 102, a RAM (Random Access Memory) 103, a communication interface 104, a display interface 105, an input interface 107, and an auxiliary storage device 120. These components are connected to a bus 110.

[0052] The control device 101 is constituted by, for example, at least one integrated circuit. The integrated circuit can be constituted by, for example, at least one CPU (Central Processing Unit), at least one GPU (Graphics Processing Unit), at least one ASIC (Application Specific Integrated Circuit), at least one FPGA (Field Programmable Gate Array), or a combination thereof.

[0053] The control device 101 controls the operation of the information processing apparatus 100 by executing various programs such as an analysis program 122 and an operating system. Based on receiving an execution instruction of various programs, the control device 101 reads the program from the auxiliary storage device 120 or the ROM 102 into the RAM 103. The RAM 103 functions as a working memory and temporarily stores various data necessary for the execution of various programs.

[0054] A LAN (Local Area Network), an antenna, etc. are connected to the communication interface 104. The information processing apparatus 100 exchanges data with external devices via the communication interface 104. The external devices include, for example, user terminals 200, servers 300, and other communication devices.

[0055] A display 106 is connected to the display interface 105. The display interface 105 sends an image signal for displaying an image to the display 106 according to a command from the control device 101 or the like. The display 106 is, for example, a liquid crystal display, an organic EL (Electro Luminescence) display, or other display device. Note that the display 106 may be integrally configured with the information processing apparatus 100 or may be configured separately from the information processing apparatus 100.

[0056] An input device 108 is connected to the input interface 107. The input device 108 is, for example, a mouse, a keyboard, a touch panel, or other device capable of receiving a user's operation. Note that the input device 108 may be integrally configured with the information processing apparatus 100 or may be configured separately from the information processing apparatus 100.

[0057] The auxiliary storage device 120 is, for example, a hard disk, a flash memory, an SSD (Solid State Drive), and other storage media. The auxiliary storage device 120 stores an analysis program 122, a patent database 124, and the above-described instruction text 126, etc. The patent database 124 includes a plurality of public information 125. The storage locations of the analysis program 122, the patent database 124, and the instruction text 126 are not limited to the auxiliary storage device 120 and may be stored in a storage area of the control device 101 (for example, a cache memory, etc.), the ROM 102, the RAM 103, an external device, or the like.

[0058] Note that the analysis program 122 may be provided by being incorporated into a part of any program, rather than as a single program. In this case, various processes defined in the analysis program 122 are realized in cooperation with any program such as the analysis program 222 described later. Even a program that does not include such a part of the module does not deviate from the gist of the analysis program 122 according to the present embodiment. Furthermore, part or all of the functions provided by the analysis program 122 may be realized by dedicated hardware. Furthermore, the information processing apparatus 100 may be configured in a form such as a so-called cloud service in which at least one server executes a part of the processing of the analysis program 122.

[0059] (C2. User terminal 200) Next, with reference to FIG. 4, the hardware configuration of the user terminal 200 shown in FIG. 1 will be described. FIG. 4 is a schematic diagram showing an example of the hardware configuration of the user terminal 200.

[0060] The user terminal 200 includes a control device 201, a ROM 202, a RAM 203, a communication interface 204, a display interface 205, an input interface 207, and an auxiliary storage device 220. These components are connected to a bus 210.

[0061] The control device 201 is constituted by, for example, at least one integrated circuit. The integrated circuit can be constituted by, for example, at least one CPU, at least one GPU, at least one ASIC, at least one FPGA, or a combination thereof.

[0062] The control device 201 controls the operation of the user terminal 200 by executing various programs such as the analysis program 222 and the operating system. Based on receiving the execution instructions of various programs, the control device 200 reads the program from the auxiliary storage device 220 or the ROM 202 into the RAM 203. The RAM 203 functions as a working memory and temporarily stores various data necessary for the execution of the program.

[0063] A LAN, an antenna, etc. are connected to the communication interface 204. The user terminal 200 exchanges data with external devices via the communication interface 204. The external devices include, for example, the information processing device 100, the server 300, and other communication devices. The user terminal 200 may be configured to be able to download the analysis program 222 from the information processing device 100.

[0064] A display 206 is connected to the display interface 205. The display interface 205 sends an image signal for displaying an image to the display 206 according to a command from the control device 201 or the like. The display 206 is, for example, a liquid crystal display, an organic EL display, or other display device. Note that the display 206 may be integrally configured with the user terminal 200 or may be configured separately from the user terminal 200.

[0065] An input device 208 is connected to the input interface 207. The input device 208 is, for example, a mouse, a keyboard, a touch panel, or other device capable of receiving a user's operation. Note that the input device 208 may be integrally configured with the user terminal 200 or may be configured separately from the user terminal 200.

[0066] The auxiliary storage device 220 is, for example, a hard disk, a flash memory, an SSD (Solid State Drive), and other storage media. The auxiliary storage device 220 stores an analysis program 222 and the like. The storage location of the analysis program 222 is not limited to the auxiliary storage device 220, and may be stored in the storage area of the control device 201 (for example, cache memory, etc.), ROM 202, RAM 203, an external device (for example, a server), or the like.

[0067] Note that the analysis program 222 may be provided not as a single program but incorporated into a part of an arbitrary program. In this case, various processes defined in the analysis program 222 are realized in cooperation with an arbitrary program such as the above-described analysis program 122. Even a program that does not include such a part of the module does not deviate from the gist of the analysis program 222 according to the present embodiment. Furthermore, some or all of the functions provided by the analysis program 222 may be realized by dedicated hardware. Furthermore, the user terminal 200 may be configured in a form such as a so-called cloud service in which at least one server executes a part of the processing of the analysis program 222.

[0068] <D. Data Flow> Next, with reference to FIGS. 5 to 13, the operation of the information processing system 10 related to the relevance evaluation of the input information 123 and the public information 125 will be described. FIG. 5 is a diagram showing an example of the data flow between the information processing apparatus 100, the user terminal 200, and the server 300.

[0069] Note that, hereinafter, the description will be made on the premise that the above-described relevance evaluation function provided by the information processing apparatus 100 is applied to the scrutiny of the public information 125. However, the application destination of the relevance evaluation function is not limited to the scrutiny of the public information 125.

[0070] (D1. Step S110) First, referring to FIG. 6, the process of step S110 shown in FIG. 5 will be described. FIG. 6 is a diagram showing an example of the setting screen 400A displayed in step S110.

[0071] The setting screen 400A accepts setting of search conditions for retrieving the publication information 125 from the above-described patent database 124 (see FIG. 3). By setting the search conditions, the user can specify one or more pieces of publication information 125 to be compared with the input information 123. The setting screen 400A includes, for example, a selection column 410 and a setting column 412.

[0072] The selection column 410 accepts selection input of the search type. As an example, the search types that can be specified in the selection column 410 include "full-text search", "field search", "command search", "semantic search", "number search", and the like.

[0073] The setting column 412 accepts input of various search conditions. Examples of the search conditions that can be input to the setting column 412 include the number of the patent document, the type of the number (for example, application number, publication number, etc.), the country of publication of the application, and the like.

[0074] Note that, in the above description, an example in which the search conditions are input by the user has been described, but the search conditions may be automatically generated based on the input information 123.

[0075] (D2. Step S112) Next, referring to FIG. 7, the process of step S112 shown in FIG. 5 will be described. FIG. 7 is a diagram showing an example of the setting screen 400B displayed in step S112.

[0076] The setting screen 400B is displayed, for example, by scrolling the above-described setting screen 400A. The setting screen 400B accepts setting of examination conditions. As an example, the setting screen 400B includes a setting column 420, a selection column 422, and a setting column 424. The setting column 424 includes an editing column 425.

[0077] The setting field 420 accepts settings for the items to be reviewed. As an example, the setting field 420 accepts the input of the maximum number of pieces of published information 125 to be reviewed and the review locations in the published information 125. Examples of the review locations that can be specified include, for example, the abstract, the title of the invention, the top claim in the claims (i.e., claim 1), the technical field, the background art, the problems to be solved by the invention, the effects of the invention, the means for solving the problems, and the specification, etc.

[0078] The selection field 422 accepts the selection of the review type. Examples of the review types that can be selected include, for example, "Investigation perspective / comparative invention", "Claim relevance evaluation", "User instructions such as summary / information extraction", and "User tag / evaluation", etc. By selecting "Claim relevance evaluation", the user can utilize the relevance evaluation function.

[0079] The setting field 424 accepts the setting of the input information 123 indicating the technical idea. The input information 123 can be set in various ways.

[0080] In a certain situation, assume that the user selects the "Claim text" button in the setting field 424. In this case, the user can input the claims by themselves in the editing field 425. For example, the user can copy & paste the claims described in some document or the claim draft created by themselves into the editing field 425. Alternatively, the user may directly input the claims into the editing field 425.

[0081] In another situation, assume that the user selects the "Publication number" button in the setting field 424. In this case, the user inputs the publication number such as the patent number or the publication number into the setting field 424 and presses the "Claim text acquisition" button. Based on this, the information processing device 100 refers to the above-mentioned patent database 124 to search for the patent document corresponding to the input publication number. Then, the information processing device 100 reflects the claims described in the searched patent document in the editing field 425.

[0082] In yet another aspect, assume that the user selects the "Technical Text" button in the setting field 424. In this case, the user designates a document indicating the technical idea. Examples of such a document include a technical paper, a newspaper article, and an invention document. Then, the user presses the "Claim Text Acquisition" button. Based on this, the information processing apparatus 100 extracts the constituent elements from the designated document by having the large language model 324 summarize the designated document.

[0083] More specifically, the information processing apparatus 100 generates an instruction sentence for extracting technical features from the input document, and inputs the instruction sentence into the large language model 324. As an example, the instruction sentence includes an instruction such as "Based on [target document], identify the technical features and generate the claims of the Japanese patent claims in Japanese." [Target document] is an argument part where the document to be summarized is designated. Thereby, the large language model 324 extracts the constituent elements from the input document. The extracted constituent elements are reflected in the editing field 425.

[0084] (D3. Steps S114, S116, S118) Next, with reference to FIGS. 8 and 9, the processing of steps S114, S116, and S118 shown in FIG. 5 will be described. FIG. 8 is a diagram showing an example of the setting screen 400C displayed in steps S114 and S118.

[0085] The setting screen 400C is displayed, for example, by scrolling the above-described setting screen 400B. As an example, the setting screen 400C includes a selection field 430, an editing field 432, a selection field 436, a selection field 438, and a start button 440. The editing field 432 includes a display field 434 for constituent elements.

[0086] The selection field 430 accepts a selection of a method for dividing the input information 123 set in the above-described editing field 425 (see FIG. 7) into constituent elements. As an example, the selection field 430 includes a "Line Break" button, an "Automated Division by AI" button, and a "Manual Input" button.

[0087] In a certain situation, assume that the user selects the "Line Break" button in the selection column 430. In this case, the information processing apparatus 100 divides the technical idea included in the input information 123 into constituent elements according to a predetermined division rule (step S116). In the example of FIG. 8, a claim, which is an example of the input information 123, is divided. The above division rule includes dividing by a predetermined keyword. Examples of the predetermined keyword include a line break code, "comprising", "and", "consisting of", etc.

[0088] In another situation, assume that the user selects the "Automatic Division by AI" button in the selection column 430. In this case, the information processing apparatus 100 divides the claims into constituent elements using the large language model 324 (step S116).

[0089] More specifically, the information processing apparatus 100 generates an instruction statement (second instruction statement) for dividing the technical idea as the input information 123 into constituent elements. FIG. 9 is a diagram showing an example of the generated instruction statement 156. The instruction statement 156 is stored in advance in the auxiliary storage device 120 of the information processing apparatus 100 as a template, for example.

[0090] The instruction statement 156 is defined to divide the input information 123 specified in the argument part 157 into constituent elements. In the argument part 157, for example, a claim, which is an example of the input information 123, is specified.

[0091] Thereafter, the information processing apparatus 100 inputs the instruction statement 156 specifying the input information 123 to the large language model 324. When receiving the input of the instruction statement 156, the large language model 324 generates an answer according to the instruction statement 156. The generated answer is output to the information processing apparatus 100. In the generated answer, the input information 123 is shown for each constituent element.

[0092] In yet another situation, assume that the user selects the "Manual Input" button in the selection column 430. In this case, the information processing apparatus 100 can freely edit the constituent elements of the claim in the editing column 432.

[0093] The input information 123 split in the way selected in the selection column 430 is displayed in the display column 434 of the editing column 432 by component. The display column 434 is displayed in the editing column 432 according to the number of split components.

[0094] As an example, when the claims input as the input information 123 are split into components #1 to #4, four display columns 434 are displayed in the editing column 432. In the example of FIG. 8, component #1 is displayed in display column 434A, component #2 is displayed in display column 434B, component #3 is displayed in display column 434C, and component #4 is displayed in display column 434D.

[0095] The user can freely edit each component displayed in the display column 434.

[0096] Also, the user can increase or decrease the number of display columns 434 displayed in the editing column 432. As an example, the user can increase the number of empty display columns 434 by pressing the "Add Component" button. Also, the user can delete the display column 434 corresponding to the button by pressing the "Delete Component" button. Furthermore, the user can delete all the display columns 434 displayed in the editing column 432 by pressing the "Delete All Components" button.

[0097] The selection column 436 accepts the selection of items to be summarized regarding the public information 125. The information processing apparatus 100 generates a summary of the public information 125 according to the items to be summarized selected in the selection column 436. The generation method will be described later.

[0098] The selection column 438 accepts the selection of the large language model 324 to be used. The large language model 324 selected in the selection column 438 is used for, among other things, the relevance evaluation between the input information 123 and the public information 125.

[0099] Based on the user pressing the review start button 440, the user terminal 200 transmits the information input on the setting screens 400A to 400C (see FIGS. 6 to 8) to the information processing apparatus 100.

[0100] In the above description, an example where the processes of steps S112, S114, S116, and S118 are executed after the process of step S110 has been described. However, the processes of steps S112, S114, S116, and S118 may be executed before the process of step S110. In this case, the search conditions in step S110 may be automatically set using the divided constituent elements #1 to #4 as the conditions for similar search. When the search conditions are automatically set, the process of step S110 does not necessarily have to be executed.

[0101] (D3. Step S120) Next, in step S120, the information processing apparatus 100 searches for the published information 125 that matches the search conditions set on the above-described setting screen 400A (see FIG. 6) from among the published information 125 registered in the above-described patent database 124. The information processing apparatus 100 uses the published information 125 that matches the search conditions as the population to be compared with the input information 123.

[0102] (D4. Step S122) Next, in step S122, the information processing apparatus 100 generates the above-described instruction text 126 (see FIG. 2) for input to the large language model 324.

[0103] FIG. 10 is a diagram showing an instruction text 126 as an example. Preferably, the instruction text 126 includes a plurality of instructions 131A to 131D.

[0104] The information processing apparatus 100 designates the input information 123 for each argument part 127 of the instructions 131A to 131D. Further, the information processing apparatus 100 designates the published information 125 for each argument part 129 of the instructions 131A to 131D.

[0105] Further, the information processing apparatus 100 sequentially specifies each of the constituent elements divided on the above-described setting screen 400C for each of the argument parts 128 of the instructions 131A to 131D.

[0106] As an example, assume that the technical idea included in the input information 123 is divided into the above-described constituent elements #1 to #4. In this case, the information processing apparatus 100 specifies the constituent element #1 for the argument part 128 of each of the instructions 131A to 131D in the argument part 128 and generates the first instruction sentence 126. Next, the information processing apparatus 100 specifies the constituent element #2 for the argument part 128 of each of the instructions 131A to 131D in the argument part 128 and generates the second instruction sentence 126. Next, the information processing apparatus 100 specifies the constituent element #3 for the argument part 128 of each of the instructions 131A to 131D in the argument part 128 and generates the third instruction sentence 126. Next, the information processing apparatus 100 specifies the constituent element #4 for the argument part 128 of each of the instructions 131A to 131D in the argument part 128 and generates the fourth instruction sentence 126. In this way, the information processing apparatus 100 repeats the generation of the instruction sentence 126 according to the number of the divided constituent elements #1 to #4.

[0107] The instruction 131A defines an instruction for extracting the description content related to each of the constituent elements #1 to #4 from the public information 125. Thereby, the user can easily grasp whether each of the constituent elements #1 to #4 is disclosed in the public information 125.

[0108] The instruction 131B defines an instruction for causing the large language model 324 to extract the description positions of the constituent elements #1 to #4 in the public information 125. The description position may be indicated by a paragraph number in a patent document, may be indicated by a claim number in a patent document, or may be indicated by the number of characters or lines from a predetermined reference position within the patent document. Thereby, the user can easily grasp the description positions of the constituent elements #1 to #4 in the public information 125.

[0109] The instruction 131C defines an instruction for causing the large language model 324 to output the reasons for extracting the description positions of the constituent elements #1 to #4 in the public information 125. In other words, the instruction 131C defines an instruction for causing the reasons for determining that each of the constituent elements #1 to #4 is described at each description position to be output. Thereby, a judgment material for determining whether each of the constituent elements #1 to #4 is disclosed in the public information 125 is provided to the user.

[0110] The instruction 131D defines an instruction for causing the large language model 324 to output the degree of relevance between each of the constituent elements #1 to #4 and the public information 125. Thereby, the user can easily grasp the degree to which each of the constituent elements #1 to #4 is relevant to the public information 125.

[0111] The information processing apparatus 100 transmits the instruction sentence 126 generated in step S122 to the server 300.

[0112] In the above description, an example in which the four instructions 131A to 131D are defined in the instruction sentence 126 has been described, but it is not necessary that all of the instructions 131A to 131D be defined in the instruction sentence 126. As an example, at least one of the instructions 131A to 131D may be defined in the instruction sentence 126.

[0113] Also, in the above description, the explanation has been made on the premise that the entire text of the public information 125 is specified in the argument part 129. However, when the amount of text of the public information 125 is large, the information processing apparatus 100 may divide the public information 125 and then compare the public information 125 with the constituent elements #1 to #4. Thereby, even when the amount of data that can be input in the usage regulations regarding the API for using the large language model 324 is limited, the user can use the API without being restricted. Also, the omission of extraction of the comparison results is suppressed.

[0114] (D5. Step S124) Next, in step S124, based on receiving the instruction text 126 from the information processing apparatus 100, the server 300 inputs the instruction text 126 into the large language model 324. As a result, the large language model 324 generates a response according to the instruction text 126.

[0115] FIG. 11 is a diagram showing response information 326 which is an example of a response generated by the large language model 324. In the example of FIG. 11, the response information 326 is shown in tabular form, but the form of the response information 326 is arbitrary. The output format of the large language model 324 is, for example, predefined in the instruction text 126, and the large language model 324 outputs the response information 326 according to the output format defined in the instruction text 126.

[0116] The response information 326 includes evaluation results 327 for each of the publicly available information 125 retrieved in step S120. In the example of FIG. 11, each of the evaluation results 327 is associated with an identifier of the publicly available information 125. The identifier of the publicly available information 125 is defined by, for example, an application number, a publication number, or a registration gazette number.

[0117] In addition, the evaluation result 327 includes, separately from the constituent elements #1 to #4 (see FIG. 8) set on the setting screen 400C, "description text", "description location", "extraction reason", and "relevance score". The "description text" is the result output from the large language model 324 according to the above-described instruction 131A (see FIG. 10). The "description location" is the result output from the large language model 324 according to the above-described instruction 131B (see FIG. 10). The "extraction reason" is the result output from the large language model 324 according to the above-described instruction 131C (see FIG. 10). The "relevance score" is the result output from the large language model 324 according to the above-described instruction 131D (see FIG. 10).

[0118] The server 300 transmits the response information 326 generated in step S124 to the information processing apparatus 100.

[0119] (D6. Step S126) Next, in step S126, the information processing apparatus 100 generates an instruction statement 166 shown in FIG. 12. FIG. 12 is a diagram showing the instruction statement 166 as an example.

[0120] The instruction statement 166 is, for example, stored in advance in the auxiliary storage device 120 of the information processing apparatus 100 as a template. The instruction statement 166 (the third instruction statement) defines an instruction for summarizing the public information 125 from a predetermined perspective.

[0121] As an example, the instruction statement 166 includes at least one of an instruction 167A for summarizing the public information 125 regarding the technical field, an instruction 167B for summarizing the public information 125 regarding the problem to be solved by the invention, an instruction 167C for summarizing the public information 125 regarding the effect of the invention, and an instruction 167D for summarizing the public information 125 regarding the operation / function.

[0122] More specifically, each of the instructions 167A to 167D includes an argument part 168. The information processing apparatus 100 designates the public information 125 to be summarized for each of the argument parts 168.

[0123] The information processing apparatus 100 transmits the instruction statement 166 generated in step S126 to the server 300.

[0124] (D7. Step S128) Next, in step S128, based on receiving the instruction statement 166 from the information processing apparatus 100, the server 300 inputs the instruction statement 166 to the large language model 324. As a result, the large language model 324 outputs a summary of the public information 125 according to the instruction statement 126.

[0125] The summary output from the large language model 324 includes a summary of the public information 125 regarding the technical field, a summary of the public information 125 regarding the problem to be solved by the invention, a summary of the public information 125 regarding the effect of the invention, and a summary of the public information 125 regarding the operation / function.

[0126] The server 300 transmits the response information generated in step S128 to the information processing apparatus 100. In this way, the information processing apparatus 100 outputs a summary of the public information 125 based on the result obtained from the large language model 324 by inputting the instruction text 166 into the large language model 324.

[0127] (D8. Steps S140, S150) Next, in step S140, the information processing apparatus 100 generates an evaluation result screen based on the response information 326 received from the server 300. The evaluation result screen is described in a language such as HTML (HyperText Markup Language).

[0128] Next, in step S150, the user terminal 200 displays the evaluation result screen generated by the information processing apparatus 100. FIG. 13 is a diagram showing an evaluation result screen 400D as an example. The evaluation result screen 400D is displayed on, for example, the display 206 of the user terminal 200.

[0129] The evaluation result screen 400D includes a display column 450 for displaying information about the input information 123, a display column 452 for displaying information about the public information 125, a display column 454 for displaying the evaluation result of the relevance between the input information 123 and the public information 125, and a display column 456 for displaying information about the summary of the public information 125.

[0130] The display column 450 displays information input to the above-described setting screens 400A, 400B (see FIGS. 6 and 7). As an example, the display column 450 displays "ID" (Identification), "Review type", "Model", "Number of acquisitions", "Status", "Acquisition time", "Review target", and "Constituent elements".

[0131] "ID" is an identifier for uniquely identifying the review evaluation result. "Review type" corresponds to the matters set in the above-mentioned selection column 422 (see Fig. 7). "Model" indicates the type of the large language model 324 used during the relevance evaluation, and corresponds to the information set in the above-mentioned selection column 438 (see Fig. 8). "Number of acquired items" corresponds to the number of pieces of public information 125 compared with the input information 123. "Status" indicates whether the relevance evaluation process has ended normally. "Acquisition time" indicates the time required from the start to the end of the execution of the relevance evaluation process. "Review target" corresponds to the information selected in the above-mentioned setting column 420 (see Fig. 7). The "constituent elements" displays the constituent elements set in the above-mentioned editing column 432 (see Fig. 8).

[0132] The display column 452 displays information related to the public information 125 that is the target of the relevance evaluation. In the display column 452, for example, "title of the invention", "solution", "selected figure", "application number", "publication number", "legal status", and "applicant" are displayed.

[0133] The display column 454 displays the evaluation result generated from the answer information 326 (see Fig. 11) of the large language model 324. As an example, as the evaluation result, for each of the constituent elements #1 to #4, the relevance score with the public information 125, the description position of the constituent elements #1 to #4 in the public information 125, the content described in the public information 125 regarding the constituent elements #1 to #4, and the determination reason for the above-mentioned relevance score are displayed. Also, the evaluation result includes the total score of the relevance of each of the constituent elements #1 to #4 and the public information 125.

[0134] The display column 456 displays a summary of the public information 125. The summary is the result obtained by inputting the above-mentioned instruction text 166 (see Fig. 12) into the large language model 324. The summary includes the summary of the public information 125 regarding the technical field, the summary of the public information 125 regarding the problem to be solved by the invention, the summary of the public information 125 regarding the effect of the invention, and the summary of the public information 125 regarding the action / function.

[0135] As described above, by extracting the comparison results between the input information 123 and the published information 125 separately for the components #1 to #4, the user can significantly reduce the time spent reading the voluminous published information 125. As a result, the user can significantly shorten the time required for invalidation searches and prior art searches.

[0136] In the example of FIG. 13, the comparison results of one piece of published information 125 are shown in the display columns 452, 454, and 456. However, in reality, the display columns 452, 454, and 456 are displayed on the evaluation result screen 400D for the number of pieces of published information 125 whose relevance has been evaluated. The evaluation results of the published information 125 are arranged, for example, in the order of the total scores of the relevance evaluations.

[0137] <E1. Others> Next, another example of the above embodiment will be described.

[0138] In the above-described setting screens 400B and 400C, an example was described in which the input information 123 is generated by dividing the claims input in text into constituent elements. However, the input information 123 indicating the technical idea does not necessarily have to be generated from text. As an example, the input information 123 indicating the technical idea may be generated from an image.

[0139] FIG. 14 is a diagram for explaining an example of generating the input information 123 from the image IM. In the example of FIG. 14, an image IM showing a PET bottle is shown.

[0140] The large language model 324 according to this example is learned to be able to interpret the content of the image. As a large language model capable of interpreting the content of an image, for example, CLIP (Contrastive Language-Image Pretraining), BLIP (Bootstrapping Language Image Pre-training for unified vision-language understanding and generation), ViLBERT (Vision-and-Language BERT), etc. can be used.

[0141] For such a large language model 324, the information processing apparatus 100 inputs the image IM and the instruction text 176. The instruction text 176 is defined to interpret the objects shown in the image IM according to the constituent elements. Thereby, the large language model 324 divides the objects shown in the image IM into constituent elements and outputs the input information 123 indicating the technical idea.

[0142] Note that the input information 123 shown in FIG. 14 is the actual output result of ChatGPT. In the output result, the plastic bottle shown in the image IM is divided into constituent elements of "bottle cap", "neck part", "body part", and "bottom part".

[0143] In this way, by extracting the technical idea from the image, even a designer who is not familiar with creating claims can easily generate the input information 123.

[0144] <E2. Others> Next, another example of the above embodiment will be described.

[0145] In the example of FIG. 10 described above, based on receiving the input of the instruction text 126, the large language model 324 outputs, as an example of the evaluation result, the positions in the public information 125 where each constituent element is described. In the above example, the paragraph numbers in the public information 125 are output as the described positions. In contrast, in this example, the drawing numbers in the public information 125 are output as the described positions.

[0146] FIG. 15 is a diagram schematically showing the process of relevance evaluation in this example. As shown in FIG. 15, the public information 125 includes one or more drawings 125F. The drawing 125F is, for example, image data. The format of the image is not particularly limited.

[0147] The information processing apparatus 100 according to this example uses the large language model 324 to compare with the input information 123 including the drawing 125F in the public information 125.

[0148] More specifically, the large language model 324 according to this example is learned to be able to interpret the content of an image. For such a large language model 324, the information processing apparatus 100 inputs the drawing 125F and a predetermined instruction sentence. The instruction sentence is defined to describe the object shown in the drawing 125F. When receiving the input of the instruction sentence, the large language model 324 outputs a description sentence of the object shown in the drawing 125F.

[0149] Thereafter, the information processing apparatus 100 instructs the large language model 324 to compare the input information 123 with the description sentence of the drawing 125F. Thereby, the information processing apparatus 100 evaluates the relevance between the input information 123 and the drawing 125F included in the public information 125. In addition, the information processing apparatus 100 also evaluates the relevance between the input information 123 and a document (for example, a specification, etc.) included in the public information 125. Since the evaluation is as described above, the description thereof will not be repeated.

[0150] Thereby, an evaluation result 130 is output. The evaluation result 130 includes not only the relevance score of each constituent element and the public information 125, but also the description position of each constituent element in the public information 125. The description position may be indicated by the paragraph number of the specification in the public information 125, or may be indicated by the figure number in the public information 125.

[0151] <E3. Others> Next, still another example of the above embodiment will be described.

[0152] In the above, it has not been particularly evaluated whether logical association of each of the public information 125 is possible. In contrast, the information processing apparatus 100 according to this example evaluates whether logical association of each of the public information 125 is possible.

[0153] FIG. 16 is a diagram showing an example of the instruction 186 used for the logical attachment evaluation. The instruction 186 is stored in advance in the auxiliary storage device 120 of the information processing apparatus 100 as a template, for example. The instruction 186 includes argument parts 187, 188, 189A, and 189B.

[0154] The instruction 186 is defined to determine whether the novelty can be negated by the combination of the patent documents specified in the argument parts 189A and 189B with respect to the input information 123 specified in the argument part 187. The input information 123 set in the argument part 187 is, for example, a claim.

[0155] A candidate for the main cited example is specified in the argument part 189A. As an example, the information processing apparatus 100 designates any of the public information 125 defined in the above-described response information 326 (see FIG. 11) in the argument part 189A. Preferably, the information processing apparatus 100 designates, in the argument part 189A, the public information 125 having the highest relevance score with the input information 123 among the public information 125 defined in the response information 326. At this time, the information processing apparatus 100 may designate the entire text of the public information 125 in the argument part 189A, or may designate a part of the public information 125 (for example, an abstract or a specification) in the argument part 189A.

[0156] A candidate for the secondary cited example is specified in the argument part 189B. As an example, the information processing apparatus 100 designates any of the public information 125 defined in the above-described response information 326 (see FIG. 11) in the argument part 189B. The document specified in the argument part 189B is different from the document specified in the argument part 189A. Preferably, the information processing apparatus 100 designates, as the secondary cited example, other public information 125 having a relevance score equal to or higher than a predetermined value (for example, 4 or higher) with respect to a constituent element whose relevance score with the main cited example is equal to or lower than a predetermined value (for example, 2 or lower) in the argument part 189B. At this time, the information processing apparatus 100 may designate the entire text of the public information 125 in the argument part 189B, or may designate a part of the public information 125 (for example, an abstract) in the argument part 189B.

[0157] Preferably, the information processing apparatus 100 designates the content of the examination criterion "Part III, Chapter 2, Section 2, Inventive Step" for the argument section 188. Thereby, the large language model 324 can determine whether it is possible to perform logical combination of the publicly available information 125 specified in the argument sections 189A and 189B after understanding the content of the examination criterion.

[0158] The embodiments disclosed this time should be considered as illustrative in all respects and not restrictive. The scope of the present invention is shown not by the above description but by the claims, and it is intended that all modifications within the meaning and scope equivalent to the claims are included.

Explanation of Reference Numerals

[0159] 10 Information processing system, 100 Information processing device, 101 Control device, 102 ROM, 103 RAM, 104 Communication interface, 105 Display interface, 106 Display, 107 Input interface, 108 Input device, 110 Bus, 120 Auxiliary storage device, 122 Analysis program, 123 Input information, 124 Patent database, 125 Public information, 125F Drawings, 126 Instruction text, 127 Argument part, 128 Argument part, 129 Argument part, 130 Evaluation result, 131A Instruction, 131B Instruction, 131C Instruction, 131D Instruction, 156 Instruction text, 157 Argument part, 166 Instruction text, 167A Instruction, 167B Instruction, 167C Instruction, 167D Instruction, 168 Argument part, 176 Instruction text, 186 Instruction text, 187 Argument part, 188 Argument part, 189A Argument part, 189B Argument part, 200 User terminal, 201 Control device, 202 ROM, 203 RAM, 204 Communication interface, 205 Display interface, 206 Display, 207 Input interface, 208 Input device, 210 Bus, 220 Auxiliary storage device, 222 Analysis program, 300 Server, 324 Large language model, 326 Answer information, 327 Evaluation result, 400A Setting screen, 400B Setting screen, 400C Setting screen, 400D Evaluation result screen, 410 Selection bar, 412 Setting bar, 420 Setting bar, 422 Selection bar, 424 Setting bar, 425 Editing bar, 430 Selection bar, 432 Editing bar, 434 Display bar, 434A Display bar, 434B Display bar, 434D Display bar, 436 Selection bar, 438 Selection bar, 440 Start button, 450 Display bar, 452 Display bar, 454 Display bar, 456 Display bar, IM Image, NW Network.

Claims

1. An analysis program, wherein the analysis program causes a computer to perform a step of obtaining input information indicating a technical idea, a step of obtaining comparison information to be compared with the technical idea, a step of dividing the technical idea into constituent elements, a step of generating a first instruction sentence including an instruction to compare each of the plurality of divided constituent elements with the comparison information, and a step of outputting an evaluation result indicating the relevance between each of the plurality of constituent elements and the comparison information based on a result obtained from the large language model by inputting the first instruction sentence into the large language model. The analysis program is as described above.

2. The dividing step includes a step of generating a second instruction sentence for dividing the technical idea into constituent elements, and a step of inputting the second instruction sentence into the large language model. The analysis program according to claim 1.

3. The dividing step includes a step of dividing the technical idea into constituent elements according to a predetermined rule. The analysis program according to claim 1.

4. The first instruction sentence includes an instruction for causing the large language model to output the description positions of the plurality of constituent elements in the comparison information. The analysis program according to any one of claims 1 to 3.

5. The comparison information includes patent documents, and the description location includes at least one of a paragraph number in the patent document, a figure number in the patent document, and a claim number in the patent document. The analysis program according to claim 4.

6. The first instruction sentence includes an instruction for causing the large language model to output the reason for extracting the description position. The analysis program according to claim 4.

7. The first instruction sentence includes an instruction for causing the large language model to output the degree of relevance between each of the plurality of constituent elements and the comparison information. The analysis program according to any one of claims 1 to 3.

8. The analysis program further causes the computer to perform a step of generating a third instruction sentence for summarizing the comparison information from a predetermined perspective, and a step of outputting a summary regarding the comparison information based on a result obtained from the large language model by inputting the third instruction sentence into the large language model. The analysis program according to any one of claims 1 to 3.

9. An information processing apparatus, comprising a control unit for controlling the information processing apparatus, the control unit performs a process of acquiring input information indicating a technical idea, a process of acquiring comparison information to be compared with the technical idea, a process of dividing the technical idea into constituent elements for each, a process of generating a first instruction sentence including an instruction to compare each of the plurality of divided constituent elements with the comparison information, and a process of outputting an evaluation result indicating the relevance between each of the plurality of constituent elements and the comparison information based on a result obtained from the large language model by inputting the first instruction sentence into the large language model. An information processing apparatus that executes.

10. An analysis method executed by a computer, comprising: a step of acquiring input information indicating a technical idea, a step of acquiring comparison information to be compared with the technical idea, a step of dividing the technical idea into constituent elements for each, a step of generating a first instruction sentence including an instruction to compare each of the plurality of divided constituent elements with the comparison information, and a step of outputting an evaluation result indicating the relevance between each of the plurality of constituent elements and the comparison information based on a result obtained from the large language model by inputting the first instruction sentence into the large language model. An analysis method.

Images