Analysis program, information processing device, and analysis method

Abstract

This analysis program for patent documents causes a computer to execute: a process of receiving input of predefined settings related to analysis of patent documents; a process of acquiring a first newly published patent document from a patent database; a process of generating an analysis result related to the first newly published patent document on the basis of a result obtained from a large language model by inputting, to the large language model, a first instruction sentence for analyzing the contents of the first newly published patent document according to the predefined settings; and a process of outputting the analysis result.

Description

Analysis Program, Information Processing Apparatus, and Analysis Method

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

[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 is trained to be able 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 the accuracy of the output is improved by inputting an instruction sentence "Let's think step by step" into ChatGPT.

[0004] "The Spell to Make ChatGPT Smarter", The Nikkei, [online], March 24, 2023, [searched on August 28, 2024], Internet <URL:https: / / www.nikkei.com / article / DGXZQOUC22BVO0S3A320C2000000 / >.

[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, it is desirable to utilize large language models for SDI (Selective Dissemination of Information) to provide new value.

[0006] In an example of the present disclosure, an analysis program for patent documents is provided. The analysis program causes a computer to execute a process of receiving a preset input related to the analysis of patent documents, a process of acquiring a first newly-arrived patent document from a patent database, a process of generating an analysis result related to the first newly-arrived patent document based on a result obtained from a large language model by inputting a first instruction sentence for analyzing the content of the first newly-arrived patent document according to the preset, and a process of outputting the analysis result.

[0007] In one example of this disclosure, the above pre-configuration includes technical information representing a technical idea. The above first instruction includes instructions for evaluating the relationship between the above technical information and the above first new patent document. The above analysis results include analysis results showing the relationship between the above technical information and the above first new patent document.

[0008] In one example of this disclosure, the output process includes the process of delivering the analysis results to the user's terminal.

[0009] In one example of this disclosure, the analysis program causes the computer to execute a process that accepts the input of filtering conditions for further narrowing down the output target. In the output process, the analysis results are output for the first newly acquired patent document that satisfies the filtering conditions.

[0010] In one example of this disclosure, the analysis program causes the computer to perform the following processes: to acquire a second new patent document that was added to the patent database after the first new patent document; to generate context information that includes some or all of the analysis results regarding the first new patent document; to input a second instruction for analyzing the content of the second new patent document into the large-scale language model and generate analysis results regarding the second new patent document based on the results obtained from the large-scale language model; and to output the analysis results regarding the second new patent document. The second instruction includes the context information.

[0011] In one example of this disclosure, the analysis program further causes the computer to execute a process to receive user evaluations regarding the analysis results for the first newly acquired patent documents. The process for generating the context information includes a process for extracting first patent documents from among the first newly acquired patent documents acquired in the acquisition process, the first patent documents whose user evaluations satisfy pre-set conditions, and a process for adding the analysis results related to the extracted first patent documents to the context information.

[0012] In one example of this disclosure, the first instruction includes an instruction to individually analyze each of the first newly acquired patent documents according to the above-mentioned pre-configuration, and an instruction to comprehensively analyze each of the first newly acquired patent documents according to the above-mentioned pre-configuration.

[0013] In one example of this disclosure, the above-mentioned pre-configuration causes the system to execute a process that accepts input of filtering conditions for narrowing down the analysis target. The above-mentioned analysis program causes the above-mentioned computer to execute a process that extracts first patent documents that satisfy the above-mentioned filtering conditions from among the first newly acquired patent documents acquired in the above-mentioned acquisition process. The above-mentioned first instruction specifies the extracted first patent documents as the analysis target.

[0014] In one example of this disclosure, the analysis program causes the computer to perform the following processes: inputting a third instruction to the large-scale language model for extracting new information that is not present in the first new patent document but newly appears in the second new patent document; and outputting the new information obtained from the large-scale language model by inputting the third instruction.

[0015] Another example of the present disclosure provides an information processing device capable of analyzing patent documents. The information processing device includes a control unit. The control unit performs the following processes: receiving input of a pre-configuration for analyzing patent documents; obtaining a first newly arrived patent document from a patent database; inputting a first instruction sentence for analyzing the content of the first newly arrived patent document in accordance with the pre-configuration into a large-scale language model, and generating an analysis result for the first newly arrived patent document based on the results obtained from the large-scale language model; and outputting the analysis result.

[0016] Another example of the present disclosure provides a method for analyzing patent documents. The analysis method comprises the steps of: receiving input of a pre-configuration for the analysis of patent documents; obtaining a first newly arrived patent document from a patent database; generating an analysis result for the first newly arrived patent document based on the results obtained from a large-scale language model by inputting a first instruction sentence for analyzing the content of the first newly arrived patent document in accordance with the pre-configuration; and outputting the analysis result.

[0017] The above and other objects, features, aspects and advantages of the present invention will become apparent from the following detailed description relating to the invention, which will be understood in conjunction with the accompanying drawings.

[0018] This figure shows an example of the device configuration of an information processing system. This figure schematically shows the SDI function according to the embodiment. This figure schematically shows the SDI function related to relevance evaluation. This is a schematic diagram showing an example of the hardware configuration of an information processing device. This is a schematic diagram showing an example of the hardware configuration of a user terminal. This figure shows an example of the data flow when performing pre-configuration. This figure shows an example of the settings screen. This figure shows an example of the data flow when performing analysis processing of newly received patent documents. This figure shows an example of an instruction statement. This figure shows an example of response information generated by a large-scale language model. This figure shows an example of a delivered email. This figure shows an example of an analysis results screen. This figure schematically shows the analysis processing according to the modified example. This figure shows a settings screen according to the modified example. This figure shows an example of an instruction statement reflecting context information. This figure shows an analysis results screen according to the modified example. This figure schematically shows the processing for extracting new information. This figure shows an instruction statement according to the modified example. This figure shows an example of an email containing the overall evaluation results. This figure explains the relevance evaluation function according to the modified example. This figure shows an example of an analysis result according to the modified example.

[0019] The embodiments of the present invention will be described below 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 of them will not be repeated. The embodiments and modifications described below may be combined selectively as appropriate.

[0020] <A. Information Processing System 10> First, the device configuration of the information processing system 10 will be explained with reference to Figure 1. Figure 1 is a diagram showing an example of the device configuration of the information processing system 10.

[0021] As shown in Figure 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 communicate with each other via a network NW (for example, the Internet).

[0022] The information processing device 100 is a notebook or desktop PC (Personal Computer), a tablet device, a smartphone, or another computer with communication capabilities. 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".

[0023] The user terminal 200 is, for example, a notebook or desktop PC, a tablet device, a smartphone, or another computer with communication capabilities. The number of user terminals 200 constituting the information processing system 10 may be one or two or more. The user terminal 200 is owned, for example, by a general user, user "A".

[0024] Server 300 is a notebook or desktop PC (Personal Computer), a tablet device, a smartphone, or another computer with communication capabilities. The number of servers 300 constituting the information processing system 10 may be one or two or more. Server 300 is operated, for example, by company "B".

[0025] Server 300 stores the Large Language Model 324. The Large Language Model 324 is a language model that has learned from a massive amount of text data, exceeding several billion entries, and is trained to process various natural languages. The Large Language Model 324 is also called an LLM (Large Language Model). The Large Language Model 324 is trained to receive an instruction as input and generate an output corresponding to that instruction.

[0026] Examples of large-scale language models 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 the known LLM. In addition to the GPT series, various other large-scale language models may be used, such as transformer-based large-scale language models like BERT (Bidirectional Encoder Representations from Transformers), CNN (Convolutional Neural Network), RNN (Recurrent Neural Network), and LSTM (Long Short-Term Memory).

[0027] Company "B" has, for example, published an API (Application Programming Interface) for utilizing the functions of the large-scale language model 324. This allows designers and general users of Company "A" to utilize the functions of the large-scale language model 324 through this API.

[0028] The various processes described herein may be implemented in the information processing device 100, the user terminal 200, the server 300, or other computers.

[0029] Furthermore, although the above description described an example in which the information processing system 10 includes a server 300, the information processing system 10 does not have to include a server 300. In this case, the information processing system 10 consists of one or more information processing devices 100 and one or more user terminals 200.

[0030] <B. Processing Overview> The information processing device 100 provides an SDI function using a large-scale language model. The SDI function is a function for monitoring the patent database and analyzing newly published patent documents. The analysis results are distributed periodically. The SDI function is used to grasp technological trends and to monitor the filing trends of competitors.

[0031] "Patent documents" refer to documents published by the patent offices of Japan or other countries. The language of patent documents may be Japanese or a foreign language such as English. Examples of patent documents include published patent gazettes, patent gazettes, published patent gazettes, republished patents, and utility model gazettes. As an example, a patent document includes bibliographic information, a specification, claims, drawings, and an abstract. Examples of bibliographic information include the application number, publication number, patent registration number, application date, publication date, registration date, applicant, patent holder, title of invention, agent, and country of application.

[0032] The following describes the SDI function provided by the information processing device 100 according to the embodiment, with reference to Figure 2. Figure 2 is a schematic diagram showing the SDI function.

[0033] The information processing device 100 receives pre-configurations 123 regarding the analysis of newly received patent documents via the user terminal 200 described above. These pre-configurations 123 include, for example, the conditions for analyzing newly received patent documents.

[0034] Furthermore, the information processing device 100 periodically monitors the patent database 124. The patent database 124 may be data stored in the information processing device 100, or it may be data stored on a server managed by the Japan Patent Office. Typically, the patent database 124 is stored in the information processing device 100. In this case, the information processing device 100 periodically monitors the Japan Patent Office's patent database and adds newly added patent documents to the patent database 124 within the information processing device 100.

[0035] The information processing device 100 retrieves newly arrived patent documents 125 to be analyzed from the patent database 124 based on the arrival of a preset analysis time. The criteria for determining whether a patent document is newly arrived can be determined by various methods. For example, the information processing device 100 considers any patent document not included in the previous analysis results as newly arrived patent documents 125. As another example, the information processing device 100 considers any patent document whose publication date belongs to the current analysis cycle as newly arrived patent documents 125.

[0036] Next, the information processing device 100 generates an instruction statement 128 (first instruction statement) for analyzing the content of the newly received patent document 125 (first newly received patent document) according to the pre-configuration 123. At this time, the instruction statement 128 reflects the pre-configuration 123 and the newly received patent document 125 to be analyzed. Subsequently, the information processing device 100 inputs the generated instruction statement 128 into the large-scale language model 324. Upon receiving the instruction statement 128, the large-scale language model 324 generates a response corresponding to the instruction statement 128. The generated response is output to the information processing device 100.

[0037] The information processing device 100 outputs the analysis results 130 of the newly received patent document 125 based on the response obtained from the large-scale language model 324. The destination of the output of the analysis results 130 is arbitrary. For example, the destination is the user terminal 200. The analysis results 130 output to the user terminal 200 are displayed on the user terminal 200's display, for example.

[0038] Furthermore, the output format of the analysis results 130 is arbitrary. For example, the analysis results 130 are delivered to the user terminal 200 in email format. As another example, the analysis results 130 are displayed on the user terminal 200 as a screen.

[0039] As described above, the information processing device 100 provides an SDI function using a large-scale language model. This allows the user to monitor the newly arrived patent documents 125 without having to set a search query.

[0040] Furthermore, the user can specify the method for analyzing the newly received patent document 125 in the pre-setting 123. For example, the user can analyze the newly received patent document 125 from the perspective of whether or not it is related to a given technical idea.

[0041] <C. Example of Use> By utilizing the SDI function provided by the information processing device 100, the user can analyze the newly received patent document 125 in various analytical modes.

[0042] As an example of the analysis method, the information processing device 100 analyzes whether a newly published patent document 125 is related to a pre-set technical idea. In this case, the user sets a technical idea to be used as a comparison standard as a pre-set 123. When a newly published patent document 125 related to the set technical idea is published, the information processing device 100 transmits the newly published patent document 125 to the user terminal 200. This allows the user to become aware that a newly published patent document 125 related to a pre-set technical idea has been published.

[0043] As another example of the analysis method, the information processing device 100 analyzes the content of the newly published patent document 125 and generates a summary. In this case, the user sets an instruction to generate a summary of the newly published patent document 125 as a pre-setting 123. As a result, the information processing device 100 generates a summary of the newly published patent document 125 and periodically transmits the generated summary to the user terminal 200. As a result, the user can simplify the process of checking the newly published patent document 125.

[0044] As yet another example of the analysis aspect, the information processing apparatus 100 analyzes the classification to which the newly arrived patent document 125 belongs from among the pre-defined classifications, and assigns the classification to the newly arrived patent document 125. Typically, the classification is a user-defined classification and is different from the patent classification. In this case, the user sets, as the pre-setting 123, an instruction to analyze the classification of the newly arrived patent document 125. Thereby, the information processing apparatus 100 assigns a classification to the newly published newly arrived patent document 125. The user can analyze the newly arrived patent document 125 based on the assigned classification. As a result, the user can shorten the time required for the confirmation work of the newly arrived patent document 125.

[0045] Hereinafter, the SDI function related to the relevance evaluation will be described, but the SDI function provided by the information processing apparatus 100 is not limited to the relevance evaluation.

[0046] FIG. 3 is a diagram schematically showing the SDI function related to the relevance evaluation. As shown in FIG. 3, the user sets, as the pre-setting 123, the technical information 123A serving as a comparison reference.

[0047] "Technical information" refers to data including a technical idea. A technical idea is a technical means for solving a technical problem. Technical information is, for example, something in which a technical means is described as a character string. Technical information includes, for example, invention information in which invention-specific matters are described.

[0048] The technical information 123A may be a character string input by the user "A" of the user terminal 200, or may be data extracted from a patent document in response to the operation of the user "A". Note that the technical information 123A may be set by specifying an application number, a publication number, or the like.

[0049] The information processing apparatus 100 acquires the newly arrived patent document 125 to be analyzed from the patent database 124 based on the arrival of the pre-set analysis time. Then, the information processing apparatus 100 performs a relevance evaluation between the pre-set technical information 123A and the newly arrived patent document 125.

[0050] The relationship between technical information 123A and the newly published patent document 125 is evaluated by utilizing a large-scale language model 324. More specifically, first, the information processing device 100 specifies technical information 123A and the newly published patent document 125 to a predetermined instruction statement 128 that is defined to evaluate the relationship between two pieces of information. The instruction statement 128 specifying technical information 123A and the newly published patent document 125 is input to the large-scale language model 324. Upon receiving the instruction statement 128, the large-scale language model 324 generates a response corresponding to the instruction statement 128. The generated response is output to the information processing device 100.

[0051] The information processing device 100 outputs an analysis result 130 showing the relationship between the technical information 123A and the newly received patent document 125, based on the response obtained from the large-scale language model 324. The analysis result 130 includes an evaluation result showing the relationship between the technical information 123A and the newly received patent document 125. The relationship may be expressed as a numerical value indicating the degree of relationship, or as an explanatory text. In the example in Figure 3, the degree of relationship is expressed as a binary value. The degree of relationship may also be expressed as a triple or greater value.

[0052] As described above, the information processing device 100 utilizes a large-scale language model 324 to evaluate the relationship between the technical information 123A and the newly received patent document 125. This allows the user to monitor the newly received patent document 125 without having to set a search query.

[0053] <D. Hardware Configuration> Next, with reference to Figures 4 and 5, the hardware configuration of the information processing device 100 and the user terminal 200 shown in Figure 1 above will be described in order.

[0054] The hardware configuration of the server 300 shown in Figure 1 is the same as that of the information processing device 100, so its explanation will not be repeated.

[0055] (D1. Information Processing Device 100) First, with reference to Figure 4, the hardware configuration of the information processing device 100 shown in Figure 1 will be described. Figure 4 is a schematic diagram showing an example of the hardware configuration of the information processing device 100.

[0056] The information processing device 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 the bus 110.

[0057] The control device 101 is comprised of, for example, at least one integrated circuit. The integrated circuit may consist of, 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.

[0058] The control device 101 controls the operation of the information processing device 100 by executing various programs such as the analysis program 122 and the operating system. Based on the receipt of execution commands for various programs, the control device 101 reads the program from the auxiliary storage device 120 or ROM 102 into the RAM 103. The RAM 103 functions as working memory and temporarily stores various data necessary for the execution of various programs.

[0059] The communication interface 104 is connected to a LAN (Local Area Network), an antenna, and other devices. The information processing device 100 exchanges data with external devices via the communication interface 104. These external devices include, for example, a user terminal 200, a server 300, and other communication devices.

[0060] A display 106 is connected to the display interface 105. The display interface 105 sends image signals to the display 106 for displaying images, in accordance with commands 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. The display 106 may be configured integrally with the information processing device 100, or it may be configured separately from the information processing device 100.

[0061] An input device 108 is connected to the input interface 107. The input device 108 is, for example, a mouse, keyboard, touch panel, or other device capable of receiving user input. The input device 108 may be configured integrally with the information processing device 100, or it may be configured separately from the information processing device 100.

[0062] The auxiliary storage device 120 is, for example, a hard disk, flash memory, SSD (Solid State Drive), or other storage medium. The auxiliary storage device 120 stores the analysis program 122, the above-mentioned pre-configuration 123, the above-mentioned patent database 124, and the above-mentioned instruction statement 128, etc. The storage location of the analysis program 122, the pre-configuration 123, the patent database 124, and the instruction statement 128 is not limited to the auxiliary storage device 120, but may also be stored in the storage area of ​​the control device 101 (for example, cache memory, etc.), ROM 102, RAM 103, external devices, etc.

[0063] Furthermore, the analysis program 122 may be provided not as a standalone program, but incorporated as part of any other program. In this case, the various processes defined in the analysis program 122 are realized in cooperation with any other program, such as the analysis program 222 described later. Even if a program does not include such modules, it does not deviate from the intent of the analysis program 122 according to this embodiment. Furthermore, some or all of the functions provided by the analysis program 122 may be realized by dedicated hardware. Moreover, the information processing device 100 may be configured in a form similar to a so-called cloud service, where at least one server executes a portion of the processing of the analysis program 122.

[0064] (D2. User Terminal 200) Next, with reference to Figure 5, the hardware configuration of the user terminal 200 shown in Figure 1 will be described. Figure 5 is a schematic diagram showing an example of the hardware configuration of the user terminal 200.

[0065] 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 the bus 210.

[0066] The control device 201 is comprised of, for example, at least one integrated circuit. The integrated circuit may consist of, for example, at least one CPU, at least one GPU, at least one ASIC, at least one FPGA, or a combination thereof.

[0067] 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 the receipt of execution commands for various programs, the control device 201 reads the program from the auxiliary storage device 220 or ROM 202 into the RAM 203. The RAM 203 functions as working memory and temporarily stores various data necessary for the execution of the program.

[0068] The communication interface 204 is connected to a LAN, an antenna, and the like. The user terminal 200 exchanges data with external devices via the communication interface 204. These external devices include, for example, an information processing device 100, a server 300, and other communication devices. The user terminal 200 may be configured to download an analysis program 222 from the information processing device 100.

[0069] A display 206 is connected to the display interface 205. The display interface 205 sends image signals to the display 206 for displaying images, in accordance with commands 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. The display 206 may be configured integrally with the user terminal 200, or it may be configured separately from the user terminal 200.

[0070] An input device 208 is connected to the input interface 207. The input device 208 is, for example, a mouse, keyboard, touch panel, or other device capable of receiving user input. The input device 208 may be configured integrally with the user terminal 200 or separately from the user terminal 200.

[0071] The auxiliary storage device 220 is, for example, a hard disk, flash memory, SSD (Solid State Drive), or other storage medium. The auxiliary storage device 220 stores the analysis program 222, etc. The storage location of the analysis program 222 is not limited to the auxiliary storage device 220, but may also be stored in the storage area of ​​the control device 201 (for example, cache memory), ROM 202, RAM 203, external devices (for example, a server), etc.

[0072] Furthermore, the analysis program 222 may be provided not as a standalone program, but incorporated as part of any other program. In this case, the various processes defined in the analysis program 222 are realized in cooperation with any other program, such as the analysis program 122 mentioned above. Even a program that does not include such modules does not deviate from the intent of the analysis program 222 according to this embodiment. Furthermore, some or all of the functions provided by the analysis program 222 may be realized by dedicated hardware. Moreover, the user terminal 200 may be configured in a form similar to a so-called cloud service, where at least one server executes part of the processing of the analysis program 222.

[0073] <E. Data Flow Related to Pre-configuration> Next, the pre-configuration performed before using the SDI function will be explained with reference to Figures 6 and 7. Figure 6 is a diagram showing an example of the data flow when performing pre-configuration.

[0074] In step S110, the user terminal 200 displays the settings screen 400A shown in Figure 7, based on the fact that it has received the pre-configured settings operation. Figure 7 is a diagram showing an example of the settings screen 400A.

[0075] The settings screen 400A is configured to accept pre-configurations for using the SDI function. The various information entered into the settings screen 400A is stored in the information processing device 100 as the pre-configurations 123 described above.

[0076] The settings screen 400A includes, for example, setting fields 410, 412, 414, 416, 418, and 420, a cancel button 430, and a save button 432.

[0077] The settings field 410 is configured to accept settings for specifying the newly received patent documents to be analyzed. For example, the settings field 410 includes radio buttons for specifying published patents as the target of analysis and radio buttons for specifying registered patents as the target of analysis. The user can select only one of the radio buttons or both radio buttons.

[0078] The settings field 412 is configured to accept settings for the analysis frequency of newly received patent documents. For example, the user can set the analysis frequency to either weekly or monthly in the settings field 412. However, the analysis frequency is not limited to weekly or monthly and may be set arbitrarily. For example, the analysis frequency may be specified by a period such as a quarter, six months, or year, or by a day of the week or date. As another example, the analysis frequency may be set immediately when the database is updated.

[0079] The settings field 414 is configured to accept various settings related to the analysis of newly submitted patent documents. For example, the settings field 414 is configured to accept input for a setting name, an instruction, the type of large-scale language model to use, the analysis location of the newly submitted patent documents, the maximum number of newly submitted patent documents to add, and the filtering conditions for the analysis results 130.

[0080] The settings section 414 includes, for example, a setting section 416 for instruction texts and a setting section 418 for filtering conditions.

[0081] The settings field 416 is configured to accept input of instruction sentences to be entered into the large-scale language model 324. The user can enter the aforementioned technical information 123A, which serves as a comparison standard, in the settings field 416. In the example in Figure 7, the technical concept, "An invention relating to the evaluation of the relevance of patent documents using a large-scale language model," is entered as text in the settings field 416.

[0082] Furthermore, the user can specify the output format of the analysis results 130 in the settings section 416. In the example in Figure 7, the user has specified that the degree of relevance between the technical information 123A and the newly received patent document 125 be output as "0" or "1". A degree of relevance of "0" indicates no relevance. A degree of relevance of "1" indicates a relevance. In addition, the user has specified in the settings section 416 that the reason for the evaluation be output.

[0083] Furthermore, various criteria can be set for determining the degree of relevance through the instructions. For example, the instructions may be set so that a relevance is determined if a specified technical term is included, and a non-relevance is determined if that term is not included. As another example, the instructions may be set so that a relevance is determined if an element is included within a specified numerical range, and a non-relevance is determined if it is outside that range or if no numerical value is provided.

[0084] The setting field 418 is configured to accept input of filtering conditions. These filtering conditions are used to narrow down the output target for the analysis results 130. When filtering conditions are set in the setting field 418, the information processing device 100 will output only the analysis results 130 that satisfy these filtering conditions.

[0085] Various filtering conditions can be set in the settings field 418. For example, the user can set a filtering condition to output only analysis results 130 that contain a specific term. As another example, the user can set a filtering condition to output only analysis results 130 whose evaluation result is a specific result. The user can arbitrarily increase or decrease the filtering conditions in the settings field 418.

[0086] The settings field 420 is configured to accept settings related to how the analysis results 130 are received. For example, the user can set in the settings field 420 whether or not to receive the analysis results 130 by email.

[0087] If the cancel button 430 is pressed, the user terminal 200 discards the information entered in the settings screen 400A and closes the settings screen 400A.

[0088] On the other hand, if the save button 432 is pressed, the user terminal 200 transmits the information entered on the settings screen 400A to the information processing device 100 as a pre-setting 123.

[0089] Referring again to Figure 6, in step S112, the information processing device 100 saves the information received from the user terminal 200 as a preset 123. The preset 123 is saved, for example, in the auxiliary storage device 120 of the information processing device 100.

[0090] <F. Data Flow Related to Analysis Processing> Next, the operation of the information processing system 10 during the analysis of the new patent document 125 will be explained with reference to Figures 8 to 12. Figure 8 is a diagram showing an example of the data flow when the analysis processing of the new patent document 125 is performed.

[0091] (F1. Step S120) In step S120, it is assumed that the analysis timing has arrived according to the analysis frequency set in the setting field 412 (see Figure 7) described above. Based on this, the information processing device 100 retrieves the new patent document 125 from the patent database 124 described above.

[0092] Preferably, the information processing device 100 extracts only newly received patent documents 125 that satisfy the filtering conditions set in the setting screen 400A (see Figure 7) as the subject of analysis. As an example, the information processing device 100 extracts newly received patent documents 125 from at least one of the published and registered publications as the subject of analysis, according to the filtering conditions set in the setting field 410 (see Figure 7) above.

[0093] The filtering criteria for narrowing down the analysis target are not limited to these. Other examples of such filtering criteria include patent classifications such as IPC, keywords, and publication dates.

[0094] In Japan, approximately several thousand publicly available patent applications are published each week, but the information processing device 100 can narrow down the target of analysis using the set filtering conditions. As a result, the cost of using the large-scale language model 324 can be reduced.

[0095] (F2. Step S122) In step S122, the information processing device 100 generates an instruction statement 128 for analyzing the contents of newly received patent documents according to the pre-settings 123. The instruction statement 128 reflects the instructions entered in the setting field 416 (see Figure 7) described above, and the newly received patent documents extracted in step S120 are designated as the target of analysis. As a result, the information processing device 100 generates an instruction statement 128 for evaluating the relationship between the technical ideas entered in the setting field 416 described above and the newly received patent documents extracted in step S120.

[0096] Figure 9 shows an example of an instruction statement 128 generated in step S122. As shown in Figure 9, the instruction statement 128 includes an instruction 132 for performing a relevance evaluation based on the technical idea entered in the setting field 416 described above, and an instruction 134 for specifying the new patent document 125 to be analyzed.

[0097] The information processing device 100 sends the instruction statement 128 generated in step S122 to the server 300.

[0098] (E3. Step S130) Next, in step S130, the server 300 inputs the instruction 128 to the large-scale language model 324 based on the fact that it has received the instruction 128 from the information processing device 100. As a result, the large-scale language model 324 generates a response corresponding to the instruction 128.

[0099] Figure 10 shows an example of response information 326 generated by the large-scale language model 324. In the example in Figure 10, the response information 326 is shown in tabular format, but the format of the response information 326 is arbitrary. The output format of the large-scale language model 324 is predetermined, for example, in the instruction statement 128, and the large-scale language model 324 outputs the response information 326 according to the output format specified in the instruction statement 128.

[0100] The response information 326 includes the analysis results 327 of the relationship between technical information 123A and new patent document 125, separately for new patent document 125. The analysis results 327 include the degree of relationship between technical information 123A and new patent document 125, and the reasons for evaluating that degree of relationship.

[0101] The analysis result 327 can be uniquely identified by the identifier of the newly received patent document 125. This identifier is defined, for example, by the application number of the newly received patent document 125, the publication number of the newly received patent document 125, the registered publication number of the newly received patent document 125, or the applicant information of the newly received patent document 125.

[0102] The information processing device 100 transmits the response information 326 generated in step S130 to the information processing device 100.

[0103] (E4. Steps S132, S140) Next, in step S132, the information processing device 100 generates an email showing the analysis results 130 of the new patent document 125 based on the response information 326 received from the server 300.

[0104] In step S140, the information processing device 100 delivers the email generated in step S132 to the user terminal 200. The user terminal 200 notifies that the email has been delivered. In this way, the information processing device 100 delivers the analysis results of newly published patent documents to the user terminal 200. By checking the email 400B, the user can determine whether or not newly published patent documents related to the technical information 123A that they have set have been published.

[0105] Figure 11 shows an example of email 400B delivered in step S140. Email 400B includes information relating to newly received patent documents, the analysis results relating to said newly received patent documents, and the reasons for the analysis.

[0106] In the example shown in Figure 11, the analysis results indicate that the degree of relevance between technical information 123A and the newly published patent document 125 is either "0" or "1". A relevance of "0" indicates that there is no relationship between technical information 123A and the newly published patent document 125. A relevance of "1" indicates that there is a relationship between technical information 123A and the newly published patent document 125. By displaying an analysis result of "0" for relevance, users can clearly understand that although the search keywords or patent classifications yielded results, there is no relevance.

[0107] Furthermore, email 400B includes message 440 to which a URL (Uniform Resource Locator) is linked. When the user clicks message 440, the user terminal 200 sends a request to the information processing device 100 to display a screen showing the details of the analysis results 130.

[0108] The means of delivering the analysis results 130 are not limited to email, but may be other means. For example, the analysis results 130 may be delivered using SMS (Short Message Service). As another example, the analysis results 130 may be delivered using a chat tool. As yet another example, the analysis results 130 may be delivered via the analysis results screen 400C (see Figure 12) described later.

[0109] (E5. Steps S142, S150) Next, in step S142, the information processing device 100 generates an analysis result screen based on the above-mentioned response information 326 (see Figure 10) received from the server 300. This analysis result screen is written in a language such as HTML (HyperText Markup Language).

[0110] Next, in step S150, the user terminal 200 displays the analysis results screen generated by the information processing device 100. Figure 12 shows an example of the analysis results screen 400C. The analysis results screen 400C is displayed, for example, on the display 206 of the user terminal 200.

[0111] The analysis results screen 400C includes a display area 450 for displaying information related to the pre-settings 123 and a display area 452 for displaying the analysis results 130.

[0112] The display area 450 displays the information set in the settings screen 400A (see Figure 7) described above. As an example, the display area 450 displays "ID" (Identification), "Analysis Type", "Model", "Status", "Number of Items Acquired", "Acquisition Time", "Analysis Target", and "User Instructions".

[0113] "ID" is an identifier used to uniquely identify the analysis results. "Analysis Type" displays the type of analysis, such as relevance evaluation. "Model" indicates the type of large-scale language model 324 used during relevance evaluation. "Number of Items Retrieved" indicates the number of newly acquired patent documents 125 that were analyzed. "Status" indicates whether the relevance evaluation process was completed successfully or not. "Acquisition Time" indicates the time taken from the start to the end of the relevance evaluation process. "Analysis Target" indicates the section of the newly acquired patent document 125 that was analyzed. "User Instructions" displays the instructions entered in the setting field 416 (see Figure 7) described above.

[0114] The display area 452 of the analysis result 130 includes display area 454 and display area 456.

[0115] The display area 454 displays information relating to the newly received patent document 125, which was the subject of the relevance evaluation. The display area 454 displays, for example, the "title of the invention," the "abstract," the "problem," the "solution," the "selected figure," the "application number," the "publication number," and the "registration number."

[0116] The display area 456 shows the analysis results generated from the response information 326 (see Figure 10) of the large-scale language model 324. As an example, the analysis results include the degree of relevance between the technical information 123A and the newly arrived patent document 125, and the reason for the evaluation that determined the degree of relevance. In the example in Figure 12, the degree of relevance is shown as "0" or "1". A degree of relevance of "0" indicates that there is no relationship between the technical information 123A and the newly arrived patent document 125. A degree of relevance of "1" indicates that there is a relationship between the technical information 123A and the newly arrived patent document 125. By displaying an analysis result of a degree of relevance of "0", the user can clearly understand that although the search keywords or patent classifications were hit, there is no relevance.

[0117] The analysis results 130 may display all of the analyzed new patent documents 125, or only those that meet the filtering conditions may be displayed. For example, filtering conditions for narrowing down the output are pre-set on the settings screen 400A (see Figure 7). Based on these filtering conditions, for example, only new patent documents 125 related to technical information 123A will be displayed on the analysis results screen 400C. This ensures that only analysis results 130 of the user's interest are displayed. As a result, the user can efficiently analyze the new patent documents 125.

[0118] Preferably, the analysis results screen 400C may be configured to display the original patent document selected by the user. This allows the user to easily verify the original patent document. When the original patent document is displayed, the analysis results related to that patent document may be displayed alongside it.

[0119] <G. Others> Next, other examples of the above embodiments will be described with reference to Figures 13 to 16.

[0120] In the above-described embodiment, past analysis trends were not utilized when analyzing the newly received patent document 125. However, the information processing device 100 may use past analysis trends to analyze the newly received patent document 125. Past analysis trends are also called contextual information. The information processing device 100 according to this example can use past analysis results 130 as contextual information to analyze the newly received patent document 125 using the same criteria as past analysis results. As a result, the user can obtain analysis results with unified judgment criteria. Consequently, the accuracy of the analysis of the newly received patent document 125 is improved.

[0121] The following describes the analysis process of newly received patent documents using contextual information, with reference to Figure 13. Figure 13 is a schematic diagram illustrating this analysis process.

[0122] In step S1, assume that the predetermined analysis time has arrived. Based on this, the information processing device 100 performs the Nth (where N is a natural number) patent analysis.

[0123] More specifically, the information processing device 100 retrieves the new patent document 125A to be analyzed from the patent database 124. Then, the information processing device 100 specifies the technical information 123A and the new patent document 125A in a predetermined instruction statement 128A for evaluating the relationship between the two pieces of information. Subsequently, the instruction statement 128A specifying the technical information 123A and the new patent document 125A is input to the large-scale language model 324. Upon receiving the instruction statement 128A, the large-scale language model 324 generates a response corresponding to the instruction statement 128A. The generated response is output to the information processing device 100.

[0124] Next, the information processing device 100 outputs an analysis result 130A showing the relationship between the technical information 123A and the newly received patent document 125A, based on the response obtained from the large-scale language model 324. The analysis result 130A includes an evaluation result showing the relationship between the technical information 123A and the newly received patent document 125A. This relationship may be expressed as a numerical value indicating the degree of relevance, or as an explanatory text.

[0125] In step S2, assume that the next analysis time has arrived. Based on this, the information processing device 100 performs the N+1th (N is a natural number) patent analysis.

[0126] More specifically, the information processing device 100 retrieves the newly arrived patent document 125B (second newly arrived patent document), which was added after the newly arrived patent document 125A, from the patent database 124. Then, the information processing device 100 specifies the technical information 123A and the newly arrived patent document 125B in a predetermined instruction statement 128B (second instruction statement) for evaluating the relationship between the two pieces of information. At this time, the information processing device 100 includes contextual information in the instruction statement 128B, including some or all of the past analysis results 130A. The instruction statement 128B includes instructions for evaluating the relationship between the technical information 123A and the newly arrived patent document 125B based on the past analysis results 130A.

[0127] Subsequently, the information processing device 100 inputs the instruction sentence 128B into the large-scale language model 324. Upon receiving the instruction sentence 128B, the large-scale language model 324 generates a response corresponding to the instruction sentence 128B. The generated response is output to the information processing device 100.

[0128] Subsequently, the information processing device 100 outputs an analysis result 130B showing the relationship between the technical information 123A and the newly received patent document 125B, based on the response obtained from the large-scale language model 324. The analysis result 130B includes an evaluation result showing the relationship between the technical information 123A and the newly received patent document 125B. This relationship may be expressed as a numerical value indicating the degree of relevance, or as an explanatory text.

[0129] Whether or not to use contextual information for analysis is predetermined in the settings screen 400D shown in Figure 14, for example. Figure 14 shows the settings screen 400D according to a modified example.

[0130] The settings screen 400D shown in Figure 14 differs from the settings screen 400A shown in Figure 7 in that it has a setting field 415 related to context information. Other points are as described above, so those explanations will not be repeated.

[0131] The settings field 415 has a checkbox that accepts whether or not to use past analysis results 130A as contextual information. If this checkbox is selected, the information processing device 100 uses the past analysis results 130A to analyze the new patent document 125B.

[0132] Figure 15 shows an example of an instruction statement 128B that reflects context information. As shown in Figure 15, the instruction statement 128B includes context information 460, 462 and instruction 464.

[0133] Context information 460 indicates the instruction text input to the large-scale language model 324 when outputting past analysis results 130A. Context information 462 shows a list of past analysis results 130A. Instruction 464 also specifies an instruction to analyze the relationship between the newly received patent document 125B and technical information 123A using context information 460 and 462.

[0134] Furthermore, if the amount of past analysis results 130A to be included in the context information increases, the cost of using the large-scale language model 324 will increase. Therefore, preferably, the setting field 415 is configured to accept settings for narrowing down the past analysis results 130A to be included in the context information.

[0135] As an example, the setting may include a setting to prioritize the inclusion of newer analysis results 130A in the context information. As another example, the setting may include a setting to specify the data size of the analysis results 130A to be included in the context information. As yet another example, the setting may include a setting to randomly select the analysis results 130A to be included in the context information. As yet another example, the setting may include a setting to select the analysis results 130A to be included in the context information based on user evaluations of past analysis results 130A. The context information may also include the reasons for the user evaluations.

[0136] An example of extracting analysis results 130A to be included in context information according to user evaluation will be further explained. In this case, the information processing device 100 has in advance received user evaluations regarding analysis results 130A for past new patent documents 125A, and extracts analysis results 130A to be included in context information based on said user evaluations.

[0137] User evaluations of the analysis results 130A are accepted in advance, for example, on the analysis results screen 400E shown in Figure 16. Figure 16 shows the analysis results screen 400E according to a modified example.

[0138] The analysis results screen 400E shown in Figure 16 differs from the analysis results screen 400C shown in Figure 12 in that it has an input field 458 for receiving user evaluations. Other points are as described above, so those explanations will not be repeated.

[0139] Input field 458 accepts user evaluations of past analysis results 130A. These user evaluations include, for example, "○" indicating that the analysis results by the large-scale language model 324 are correct, and "×" indicating that the analysis results by the large-scale language model 324 are incorrect. The user evaluations entered in input field 458 are associated with each of the newly received patent documents 125A and stored in the information processing device 100.

[0140] When the information processing device 100 analyzes the newly received patent document 125B, it extracts from past newly received patent documents 125A that meet pre-set user evaluation criteria. For example, the information processing device 100 extracts newly received patent documents 125A with a user evaluation of "○". The information processing device 100 then includes the analysis results 130A related to the extracted newly received patent documents 125A in the context information. As a result, only the analysis results 130A that match the user evaluation are included in the context information. Consequently, the accuracy of subsequent analyses related to newly received patent documents 125B is improved.

[0141] <H. Others> Next, with reference to Figure 17, yet another example of the above embodiment will be described.

[0142] In the above example, the information processing device 100 used past analysis results 130A as contextual information to analyze the newly received patent document 125B. In contrast, the information processing device 100 according to this example uses a large-scale language model 324 to extract new information that was not present in the past patent document 125A but newly appeared in the current patent document 125B.

[0143] Examples of such new information include new technical terms, new challenges, new applicants, and patent classifications. Patent classifications include IPC (International Patent Classification), FI (File Index), and F-terms (File Forming Term).

[0144] The extracted new information can be used for various purposes. For example, the extracted new information can be used as search criteria for newly published patent documents. This new information may be automatically reflected in the search criteria or manually. By using the new information in the search criteria, the setting of search criteria is simplified. Furthermore, by reviewing the extracted new information, users can identify issues that have not appeared in previous analysis results by comparing them. In addition, because users can review the extracted new information, they can identify rapidly increasing technological fields and patent classifications.

[0145] A sudden influx of new information may be highlighted more than other new information. The method of highlighting is optional. For example, a sudden influx of new information can be highlighted by displaying it in a specific color. Another example is highlighting a sudden influx of new information by adding a line break. Yet another example is highlighting a sudden influx of new information by enclosing it in quotation marks. Yet another example is highlighting a sudden influx of new information by underlining it. Yet another example is highlighting a sudden influx of new information by making it bold. Yet another example is highlighting a sudden influx of new information by adding a callout.

[0146] Figure 17 is a schematic diagram illustrating the process for extracting new information. As described above, the newly acquired patent document 125A is the data used when outputting the Nth analysis result 130A. The newly acquired patent document 125B is the data used when outputting the N+1th analysis result 130B.

[0147] The information processing device 100 generates an instruction statement 128C (third instruction statement) for extracting new information that has newly appeared in the new patent document 125B, based on the previously received new patent document 125A. The instruction statement 128C includes, for example, an instruction for extracting new technical words related to the technical field of the large-scale language model. The information processing device 100 outputs new information 131 obtained from the large-scale language model 324 by inputting the instruction statement 128C into the large-scale language model 324.

[0148] The destination for outputting the new information 131 is arbitrary. For example, the destination is the user terminal 200. The new information 131 output to the user terminal 200 is displayed on the user terminal 200's screen, for example. The output method of the new information 131 is also arbitrary. For example, the new information 131 is delivered to the user terminal 200 in email format. As another example, the new information 131 is displayed on the user terminal 200 in screen format.

[0149] The user can select any information from the displayed new information 131. The information processing device 100 uses the information selected by the user when analyzing the next new patent document 125.

[0150] <I. Others> Next, with reference to Figures 18 and 19, yet another example of the above embodiment will be described.

[0151] In the above-described embodiment, the information processing device 100 analyzed each of the newly acquired patent documents 125 individually. In contrast, the information processing device 100 may analyze each of the newly acquired patent documents 125 comprehensively.

[0152] Figure 18 shows an instruction statement 128D that follows a modified example. The instruction statement 128D shown in Figure 18 differs from the instruction statement 128 described in Figure 9 above in that it further includes instruction 133.

[0153] More specifically, instruction 128D includes not only instruction 132 for individually analyzing each of the new patent documents 125 according to the pre-configuration 123, but also instruction 133 for comprehensively analyzing each of the new patent documents 125 according to the pre-configuration 123. The information processing device 100 inputs instruction 128D into the large-scale language model 324 to obtain individual analysis results for the new patent documents 125 and a comprehensive analysis result for the new patent documents 125.

[0154] The comprehensive analysis results are delivered, for example, as email 400B, as shown in Figure 19. Figure 19 is a diagram showing an example of email 400B containing the comprehensive evaluation results. By checking email 400B, users can understand new applicants, technological trends, and more.

[0155] Preferably, the overall analysis results are displayed above the individual analysis results of the newly filed patent document 125. This allows the user to first review the overall summary of the analysis results and easily determine whether or not the results are important.

[0156] In the above description, we explained an example where the comprehensive analysis results are included in email 400B, but the comprehensive analysis results may also be displayed on the analysis results screen 400C (see Figure 12) as described above.

[0157] <J. Others> Next, with reference to Figures 20 and 21, yet another example of the above embodiment will be described.

[0158] In the above-described embodiment, the information processing device 100 evaluated the relationship between the technical information 123A and the newly published patent document 125. In this case, the information processing device 100 may also evaluate the relationship between the technical information 123A and the newly published patent document 125 from a specific viewpoint.

[0159] Figure 20 is a diagram illustrating the relationship evaluation function according to this example. As shown in Figure 20, the information processing device 100 specifies the above-mentioned technical information 123A and the above-mentioned new patent document 125 as a predetermined instruction statement 128T for evaluating the relationship between two pieces of information from a specific viewpoint. The instruction statement 128T is pre-registered in the information processing device 100 as a template, for example.

[0160] More specifically, instruction statement 128T includes argument section 129A and argument section 129B. Argument section 129A specifies technical information 123A. On the other hand, argument section 129B specifies newly arrived patent document 125.

[0161] Furthermore, instruction 128T specifies the criteria from which the relationship between technical information 123A and newly received patent document 125 will be evaluated. In the example in Figure 20, instruction 128T specifies that two pieces of information should be evaluated for each of the multiple evaluation criteria "α", "β", ... Each evaluation criterion may be predetermined or may be defined as an argument. If each evaluation criterion is defined as an argument, that evaluation criterion is predetermined in the setting screen 400A described above. Note that instruction 128T may specify only one evaluation criterion.

[0162] The evaluation criteria include at least one of the following: commonality in the technical field, commonality in the problem, commonality in the function, commonality in the content suggestion, commonality in the parameters, commonality in the numerical limitations, and commonality in the technical terminology. This allows the commonality between technical information 123A and newly received patent document 125 to be evaluated using an evaluation axis that aligns with the criteria for determining inventive step.

[0163] As another example, the evaluation criteria may be criteria suitable for application exploration. These evaluation criteria may be criteria for commonality of applicable applications or products, or criteria for commonality of applicable industrial fields. Yet another example is that the evaluation criteria may be criteria for commonality of drawings.

[0164] Instruction 128, which specifies technical information 123A and new patent document 125, is input to the large-scale language model 324. Upon receiving instruction 128, the large-scale language model 324 generates a response corresponding to instruction 128. The generated response is output to the information processing device 100.

[0165] The information processing device 100 outputs an analysis result 130C that shows the relationship between the technical information 123A and the newly received patent document 125 for each evaluation criterion, based on the response obtained from the large-scale language model 324. This relationship may be expressed as a numerical value indicating the degree of relevance, or as an explanatory text.

[0166] Figure 21 shows an example of the analysis result 130C in this example. In the example in Figure 21, the analysis result 130C is shown in tabular format, but the format of the analysis result 130C is arbitrary. The output format of the large-scale language model 324 is predetermined in, for example, instruction 128, and the large-scale language model 324 outputs the analysis result 130C according to the output format specified in instruction 128.

[0167] Analysis result 130C includes the degree of relevance between technical information 123A and newly received patent document 125, and the reasons for evaluating that degree of relevance. The degree of relevance is included for each evaluation criterion specified in instruction 128. Similarly, the reasons for evaluation are included for each evaluation criterion specified in instruction 128. Furthermore, analysis result 130C includes the overall evaluation score for each evaluation criterion and the reasons for that overall evaluation.

[0168] The output destination of the analysis results 130C from the information processing device 100 is arbitrary. For example, the output destination is the user terminal 200. The analysis results 130C output to the user terminal 200 are displayed on the user terminal 200's display, for example.

[0169] The embodiments disclosed herein should be considered in all respects to be illustrative and not restrictive. The scope of the invention is indicated by the claims rather than by the foregoing description, and all modifications within the meaning and scope of equivalents of the claims are intended.

[0170] 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 Pre-configuration, 123A Technical information, 124 Patent database, 125 New patent documents, 125A New patent documents, 125B New patent documents, 128 Instructions, 128A Instructions, 128B Instructions, 128C Instructions, 128D Instructions, 128T Instructions, 129A Argument section, 129B Argument section, 130 Analysis results, 130A Analysis results, 130B Analysis results, 130C Analysis results, 131 New information, 132 Instructions, 133 Instructions, 134 Instructions, 200 User terminal, 201 Control unit, 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-scale language model, 326 Response information, 327 Analysis results, 400A Settings screen, 400B Email, 400C Analysis results screen, 400D Settings screen, 400E Analysis results screen, 410 Settings field, 412 Settings field, 414 Settings field, 415 Settings field, 416 Settings field, 418 Settings field, 420 Settings field, 430 Cancel button, 432 Save button, 440 Message, 450 Display field, 452 Display field, 454 Display field, 456 Display field, 458 Input field, 460 Context information, 462 Context information, 464 Instructions.

Claims

1. A patent document analysis program, wherein the analysis program causes a computer to perform the following processes: a process of receiving input for pre-configuration regarding the analysis of patent documents; a process of obtaining a first newly arrived patent document from a patent database; a process of generating an analysis result regarding the first newly arrived patent document based on the results obtained from a large-scale language model by inputting a first instruction sentence for analyzing the content of the first newly arrived patent document according to the pre-configuration; and a process of outputting the analysis result.

2. The analysis program according to claim 1, wherein the pre-configuration includes technical information representing a technical idea, the first instruction statement includes instructions for evaluating the relationship between the technical information and the first newly filed patent document, and the analysis result includes an analysis result showing the relationship between the technical information and the first newly filed patent document.

3. The analysis program according to claim 1 or 2, wherein the output process includes a process for delivering the analysis results to a user terminal.

4. The analysis program according to claim 1 or 2, wherein the analysis program causes the computer to execute a process to accept input of filtering conditions for narrowing down the output target, and in the output process, the analysis results are output for the first newly acquired patent document that satisfies the filtering conditions.

5. The analysis program further causes the computer to perform the following processes: acquiring a second new patent document that was added to the patent database after the first new patent document; generating context information that includes part or all of the analysis results relating to the first new patent document; inputting a second instruction for analyzing the content of the second new patent document into the large-scale language model and generating analysis results relating to the second new patent document based on the results obtained from the large-scale language model; and outputting the analysis results relating to the second new patent document, wherein the second instruction includes the context information, the analysis program according to claim 4.

6. The analysis program according to claim 5, wherein the analysis program further causes the computer to perform a process to receive user evaluations regarding the analysis results for the first newly acquired patent documents, and the process for generating the context information includes a process for extracting first patent documents from among the first newly acquired patent documents acquired in the acquisition process for which the user evaluations satisfy pre-set conditions, and a process for adding the analysis results relating to the extracted first patent documents to the context information.

7. The analysis program according to claim 1 or 2, wherein the first instruction includes an instruction to individually analyze each of the first newly acquired patent documents according to the pre-configuration, and an instruction to comprehensively analyze each of the first newly acquired patent documents according to the pre-configuration.

8. The analysis program according to claim 1 or 2, wherein the pre-configuration causes the computer to execute a process that accepts input of filtering conditions for narrowing down the analysis target, the analysis program further causes the computer to execute a process that extracts first patent documents that satisfy the filtering conditions from among the first newly acquired patent documents acquired in the acquisition process, and the first instruction statement specifies the extracted first patent documents as the analysis target.

9. The analysis program according to claim 5, wherein the analysis program causes the computer to further input a third instruction to the large-scale language model for extracting new information that is not in the first new patent document but newly appears in the second new patent document, and outputs the new information obtained from the large-scale language model by inputting the third instruction.

10. An information processing device capable of analyzing patent documents, comprising a control unit, the control unit performing the following processes: receiving input for pre-settings related to the analysis of patent documents; obtaining a first newly arrived patent document from a patent database; inputting a first instruction sentence for analyzing the content of the first newly arrived patent document in accordance with the pre-settings into a large-scale language model, and generating analysis results related to the first newly arrived patent document based on the results obtained from the large-scale language model; and outputting the analysis results.

11. A method for analyzing patent documents, comprising: a step of receiving input for pre-configuration regarding the analysis of patent documents; a step of obtaining a first newly arrived patent document from a patent database; a step of inputting a first instruction sentence for analyzing the content of the first newly arrived patent document in accordance with the pre-configuration into a large-scale language model, and generating an analysis result regarding the first newly arrived patent document based on the results obtained from the large-scale language model; and a step of outputting the analysis result.

Images