Program, method, information processing apparatus

Abstract

Provided are a program, a method, and an information processing apparatus for operating a computer including a processor that can further reduce the burden on an expert for recording the consultation content of a client. 【Solution means】In a system, a program of a server 20 of a case management service for managing legal consultation cases causes a computer processor to execute steps of: receiving an input of consultation data indicating consultation content from a client; obtaining a result of summarizing the consultation content by giving a prompt to a large language model so as to summarize the consultation content indicated by the consultation data along predetermined items; and associating the consultation content summarized along the predetermined items with the client and storing the result in a storage unit so that an expert responding to the consultation from the client can refer to it.

Description

Technical Field

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

Background Art

[0002] Various stakeholders, such as ordinary consumers and businesses, are conducting procedures based on laws. Therefore, there are many opportunities to consult legal experts such as lawyers.

[0003] Patent Document 1 describes a technology that, with the problem of "providing legal consultation services at low cost", "connects the providers and recipients of legal consultation services using an electronic network and provides legal consultation services via the electronic network".

[0004] According to the technology of Patent Document 1, it is possible to "control the transmission and reception of legal consultation service information between the provider terminal and the recipient terminal using an electronic network", and it is said to have the effect of "being able to provide legal consultation services at a lower cost".

Prior Art Documents

Patent Documents

[0005]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0006] When an expert interviews a client about the content of the consultation and creates legal documents such as contracts, warning letters, pleadings, and answers, it is necessary to organize the content of the consultation. However, if recording the content of the consultation becomes a burden, it may be postponed. As a result, it may be necessary to listen to the recording again to recall the content of the consultation, which may actually require more man-hours.

[0007] Therefore, there is a need for a technology that can further reduce the burden on experts to record the consultation content of the consulters.

Means for Solving the Problem

[0008] According to an embodiment shown in the present disclosure, a program for operating a computer including a computer processor is provided. The program causes the computer processor to execute steps of receiving an input of consultation data indicating the consultation content from the consulter, obtaining a result of summarizing the consultation content by giving a prompt to a large language model so as to summarize the consultation content along predetermined items, and associating the consultation content summarized along the predetermined items with the consulter and storing it in a storage unit so that an expert responding to the consultation from the consulter can refer to it.

Advantages of the Invention

[0009] According to the present disclosure, the burden on experts to record the consultation content of the consulters can be further reduced.

Brief Description of the Drawings

[0010]

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

Figure 17

Figure 18

Figure 19

Figure 20

Figure 21

Figure 22

Figure 23

Figure 24

Figure 25

Figure 26

Figure 27

Figure 28

Figure 29

Figure 30

Mode for Carrying Out the Invention

[0011] Hereinafter, embodiments of the present disclosure will be described with reference to the drawings. In the following description, the same parts are denoted by the same reference numerals. Their names and functions are also the same. Therefore, detailed descriptions thereof will not be repeated.

[0012] <Outline of Embodiment> <1.1 Configuration Diagram of the Whole System> FIG. 1 is a diagram showing the configuration of system 1.

[0013] The system 1 shown in FIG. 1 includes a server 20 for a case management service that manages legal consultation cases, a user terminal 10, a server 91 for a case law search service, a server 92 for a statute search service, a server 93 for a book viewing service, a server 94 for an information media service, a server 95 for an artificial intelligence (large language model) service (hereinafter sometimes referred to as the "server 95 for large language model service"), a server 96 of a business operator, a server 97 of an SNS, and a user terminal 10A. These devices are communicatively connected via a network 80.

[0014] In the illustrated example, the terminals used by the users of the case management service provided by the server 20 are shown as the terminal 10, the terminal 10A, etc., but each user is operating a terminal.

[0015] In the present embodiment, each device (terminal device, server, etc.) can also be regarded as an information processing device. That is, the aggregate of each device can be regarded as one "information processing device", and the system 1 may be formed as an aggregate of a plurality of devices. The way of distributing the plurality of functions required to implement the system 1 according to the present embodiment for one or more pieces of hardware can be appropriately determined in view of the processing capabilities of each piece of hardware and / or the specifications required for the system 1.

[0016] The terminal 10 is a device operated by the user. In the present embodiment, a user who is an expert in responding to legal consultations operates the terminal 10. The expert user creates legal documents such as contracts, warning letters, response letters, pleadings, and pleadings according to the consultation content. The terminal 10 and the terminal 10A have the same functional configuration. The terminal 10 is realized, for example, as follows. · A handheld mobile terminal such as a smartphone or a tablet · A desktop PC (Personal Computer) or a laptop PC · A wearable terminal (such as a wristwatch type or a glasses type) worn by the user The terminal 10 includes a communication IF (Interface) 12, an input device 13, an output device 14, a memory 15, a storage 16, and a processor 19.

[0017] The communication IF 12 is an interface for inputting and outputting signals so that the terminal 10 can communicate with an external device.

[0018] The input device 13 is a device for receiving input operations from a user (for example, a pointing device such as a touch panel, a touch pad, a mouse, etc., a keyboard, etc.).

[0019] The output device 14 is a device for presenting information to the user (a display, a speaker, etc.).

[0020] The memory 15 is for temporarily storing programs and data processed by programs, etc., and is a volatile memory such as a DRAM (Dynamic Random Access Memory), for example.

[0021] The storage 16 is for storing data, and is, for example, a flash memory or an HDD (Hard Disc Drive).

[0022] The processor 19 is hardware for executing an instruction set described in a program, and is composed of an arithmetic unit, registers, peripheral circuits, etc.

[0023] Server 20 is a device for providing users with a service for managing cases that are units of legal consultation services. In this embodiment, Server 20 provides functions for receiving registration of information related to a case from a user of Terminal 10, while also summarizing the consultation content, investigating the arguments related to the consultation content, investigating precedents or books corresponding to the arguments, and creating legal documents. Based on the case information, Server 20 causes the server 95 of the artificial intelligence (large language model) service to generate information related to these functions, and provides a case management service by recording the generation results in Server 20 while responding to the user.

[0024] Also, in this embodiment, Server 20 is also a device for providing users with legal consultation services. In this embodiment, Server 20 receives input of legal consultations in free text in a chat format from a user of Terminal 10, causes the server 95 of the artificial intelligence (large language model) service to generate an answer to the received legal consultation content, and provides a legal consultation service by responding to the user with the generation result. As users, there can be consultants who conduct consultations and experts who respond to consultations. Specifically, Server 20 provides legal consultation services to the following types of users. · Users who perform legal work and answer legal consultations, such as the legal departments of business companies, etc. · Users who do not necessarily specialize in legal work, such as business departments of business companies, etc. · Users who provide legal consultations as a service to clients as experts, such as law firms, etc. Server 20 may also accept requests for legal consultations from clients to experts by matching users who conduct legal consultation services as experts, such as law firms, etc., with users who request legal consultations from experts such as business companies or individuals. For example, it may be in a form where the consultation content of the consultant is reflected on a bulletin board viewable by third parties and the expert's answer is also made public, or it may be in a form where the consultation content of the consultant is not disclosed to third parties and is in a non-public state, allowing consultations with the expert. Server 20 accumulates such consultation content of clients and answer content of experts.

[0025] Server 20 includes a communication IF 22, an input / output IF 23, a memory 25, a storage 26, and a processor 29.

[0026] The communication IF 22 is an interface for inputting and outputting signals so that the server 20 can communicate with an external device.

[0027] The input / output IF 23 functions as an interface with an input device for receiving input operations from a user and an output device for presenting information to the user.

[0028] The memory 25 is for temporarily storing programs and data processed by programs and the like, and is a volatile memory such as a DRAM (Dynamic Random Access Memory).

[0029] The storage 26 is for storing data, such as a flash memory or an HDD (Hard Disc Drive).

[0030] The processor 29 is hardware for executing an instruction set described in a program, and is composed of an arithmetic unit, registers, peripheral circuits, and the like.

[0031] The server 91 of the case search service has a case database and enables a user to search for judgments determined by a court or the like.

[0032] The server 92 of the statute search service has a statute database and enables a user to search for statutes.

[0033] The server 93 of the book viewing service is a service that enables viewing of electronic contents such as magazines and books. For example, by a user periodically paying a fixed fee, electronic books in the legal field or the like can be viewed.

[0034] The server 94 of the information media service provides a service for providing information, such as a service that enables browsing by collecting blog posts, Q&A sites, news articles, IR information, etc. The server 94 of the information media service accumulates information on interpretations based on laws and regulations and guidelines of public offices, etc. prepared to inform operation rules.

[0035] In addition to these, the server 94 of the information media service may also provide a service for collecting and providing the following information. · Whether it is a group that meets the conditions for contract cancellation, such as belonging to an antisocial force. For example, there may be a database of business operators, etc. that meet the above conditions, or public information such as news articles. · The reputation of a group such as a business company, etc., where the customer has a negative reaction and it has been picked up and spread on news, SNS, etc. For example, on SNS, etc., information may have spread with a negative keyword and a certain number of impressions or more. · Information related to performance, such as sales, revenue (amount of profit, profit rate, etc.), evaluation results related to business sustainability (continuing deficit, bankruptcy possibility, bond rating, etc.). The server 95 of the large language model service is a server that executes language processing tasks using a language model constructed by learning processing including artificial intelligence (AI). An LLM (Large Language Model) is one that has learned a large amount of data (text data, etc.) in advance, such as Web content on the Internet, or data accumulated in a predetermined database, and can execute various language processing tasks by being given a task.

[0036] The server 95 of the large language model service accepts input of prompts by text, image, voice, etc., generates an answer to the prompt, and responds. Examples of LLM include GPT-3 and GPT-4 developed by OpenAI, BERT developed by Google, etc.

[0037] The operator's server 96 is for the operator to conduct business activities and store data generated in the course of business. These data have viewing permissions set, and whether viewing is allowed is set according to the data for users belonging to the operator and external users who do not belong to the operator.

[0038] The SNS server 97 provides a service that promotes communication between users. For example, it is a service that allows each user to view the content posted by other users. For example, there are services that can be viewed through Internet searches even without having a user account on the SNS, and services that can view the posts of each user when having a user account.

[0039] <1.2 Functional Configuration of Server 20> Figure 2 is a diagram showing the configuration of server 20. As shown in Figure 2, server 20 functions as a communication unit 201, a storage unit 202, and a control unit 203.

[0040] The communication unit 201 performs processing for the server 20 to communicate with external devices.

[0041] The storage unit 202 stores various databases such as a user database 211, a chat consultation history database 212, an LLM usage history database 213, a prompt database 214, an expert consultation history database 215, and a project management database 216.

[0042] The user database 211 is a database that manages information of each user. Details will be described later.

[0043] The chat consultation history database 212 is a database that shows the history of users consulting using the services provided by the server 20. Details will be described later.

[0044] The LLM utilization history database 213 is a database that shows the history of answers generated by the server 95 of the artificial intelligence service for the consultation content of the user. Details will be described later.

[0045] The prompt database 214 is a database that manages the templates of prompts to be sent to the server 95 of the large language model service. Details will be described later.

[0046] The expert consultation history database 215 is a database that shows the history of consultations made by the user, who is the consultee, to the expert. Details will be described later.

[0047] The case management database 216 is a database that manages the cases that are the units of the business related to the consultation. Details will be described later.

[0048] The control unit 203 is realized by the processor 29 reading the program stored in the storage unit 202 and executing the instructions included in the program. By operating according to the program, the control unit 203 exhibits the functions shown as the reception control module 2041, the transmission control module 2042, the user management module 2043, the chat consultation processing module 2044, the LLM utilization module 2045, the learning processing module 2046, the matching processing module 2047, and the case management module 2048.

[0049] The reception control module 2041 controls the process of the server 20 receiving signals from an external device according to the communication protocol.

[0050] The transmission control module 2042 controls the process of the server 20 transmitting signals to an external device according to the communication protocol.

[0051] The user management module 2043 is a module for managing the information of each user who uses the system 1. Specifically, the user management module 2043 accepts the registration of the information of each user and updates the user database 211.

[0052] The chat consultation processing module 2044 is a program module that receives and responds to legal consultation inputs from users via chat.

[0053] Specifically, the chat consultation processing module 2044 receives the input of the consultation content from the user, causes the server 95 of the large language model service to generate an answer, and updates the chat consultation history database 212.

[0054] The LLM utilization module 2045 is a program module that generates a prompt based on the content of the consultation input by the user in free text, image, voice, etc., and generates an answer by sending the generated prompt to the server 95 of the large language model service.

[0055] Specifically, the LLM utilization module 2045 manages the text input by the user as a series, and updates the LLM utilization history database 213 based on the prompt sent to the server 95 of the large language model service and the response from the server 95 of the large language model service.

[0056] The learning processing module 2046 is a program module that trains a learned model that outputs an answer to a question based on the data accumulated as the chat consultation history database 212, etc. on the server 20.

[0057] The matching processing module 2047 is a program module that receives an operation of entrusting a legal consultation from a counselor to an expert and matches with an expert such as a lawyer.

[0058] The case management module 2048 is a program module that receives the registration of various information about a case, records it in a record, reads out the information recorded in the record, and presents it to the user.

[0059] The case management module 2048 gives a deadline-based notification when a deadline is set for a case.

[0060] <1.3 Configuration of the terminal 10> FIG. 3 is a diagram showing the configuration of the terminal 10.

[0061] As shown in FIG. 3, the terminal 10 includes a plurality of antennas (antenna 111, antenna 112), a communication unit corresponding to each antenna (first communication unit 120, second communication unit 121), an input device 130 (including a touch-sensitive device 131), a display 132, an audio processing unit 140, a microphone 141, a speaker 142, a position information sensor 150, a camera 160, a motion sensor 170, a storage unit 180, and a control unit 190. The terminal 10 also has functions and configurations not particularly shown in FIG. 3 (for example, a battery for holding power, a power supply circuit for controlling the supply of power from the battery to each circuit, etc.). As shown in FIG. 3, each block included in the terminal 10 is electrically connected by a bus or the like.

[0062] The antenna 111 radiates the signal emitted by the terminal 10 as a radio wave. Also, the antenna 111 receives a radio wave from space and gives the received signal to the first communication unit 120.

[0063] The antenna 112 radiates the signal emitted by the terminal 10 as a radio wave. Also, the antenna 112 receives a radio wave from space and gives the received signal to the second communication unit 121.

[0064] The first communication unit 120 performs modulation / demodulation processing and the like for transmitting and receiving signals via the antenna 111 in order for the terminal 10 to communicate with other wireless devices. The second communication unit 121 performs modulation / demodulation processing and the like for transmitting and receiving signals via the antenna 112 in order for the terminal 10 to communicate with other wireless devices. The first communication unit 120 and the second communication unit 121 are communication modules including a tuner, an RSSI (Received Signal Strength Indicator) calculation circuit, a CRC (Cyclic Redundancy Check) calculation circuit, a high-frequency circuit, and the like. The first communication unit 120 and the second communication unit 121 perform modulation / demodulation, frequency conversion, etc. of the wireless signals transmitted and received by the terminal 10, and supply the received signals to the control unit 190.

[0065] The input device 130 has a mechanism for receiving a user's input operation. Specifically, the input device 130 is configured as a touch screen and includes a touch-sensitive device 131. The touch-sensitive device 131 receives a user's input operation on the terminal 10. The touch-sensitive device 131 detects the contact position of the user on the touch panel, for example, by using a capacitive touch panel. The touch-sensitive device 131 outputs a signal indicating the contact position of the user detected by the touch panel to the control unit 190 as an input operation.

[0066] The display 132 displays data such as images, videos, and texts according to the control of the control unit 190. The display 132 is realized by, for example, an LCD, an organic EL display, or the like.

[0067] The audio processing unit 140 demodulates and modulates an audio signal. The audio processing unit 140 modulates the signal given from the microphone 141 and gives the modulated signal to the control unit 190. Also, the audio processing unit 140 gives the audio signal to the speaker 142. The audio processing unit 140 is realized by, for example, a processor for audio processing. The microphone 141 receives an audio input and gives an audio signal corresponding to the audio input to the audio processing unit 140. The speaker 142 converts the audio signal given from the audio processing unit 140 into audio and outputs the audio to the outside of the terminal 10.

[0068] The position information sensor 150 is a sensor that detects the position of the terminal 10, and is, for example, a GPS (Global Positioning System) module. The GPS module is a receiving device used in a satellite positioning system. In the satellite positioning system, signals from at least three or four satellites are received, and based on the received signals, the current position of the terminal 10 equipped with the GPS module is detected.

[0069] The camera 160 is a device that receives light by a light receiving element and outputs it as a captured image. The camera 160 is, for example, a depth camera that can detect the distance from the camera 160 to the object to be photographed.

[0070] The motion sensor 170 includes an acceleration sensor, an angular velocity sensor, etc., and detects the movement of the terminal 10.

[0071] The storage unit 180 is composed of, for example, a flash memory or the like, and stores data and programs used by the terminal 10. Various information stored in the storage unit 180 will be described later.

[0072] The control unit 190 controls the operation of the terminal 10 by reading a program stored in the storage unit 180 and executing instructions included in the program. The control unit 190 is, for example, an application processor. By operating according to the program, the control unit 190 functions as an operation reception unit 191, a transmission / reception unit 192, a data processing unit 193, a notification control unit 194, and a storage control unit 195.

[0073] The operation reception unit 191 performs a process of receiving a user's input operation on an input device such as the touch-sensitive device 131. Based on the information of the coordinates where the user touches the touch-sensitive device 131 with a finger or the like, the operation reception unit 191 determines the type of operation, such as whether the user's operation is a flick operation, a tap operation, a drag (swipe) operation, or the like.

[0074] The transmission / reception unit 192 performs a process for the terminal 10 to transmit and receive data to and from an external device such as the server 20 according to a communication protocol.

[0075] The data processing unit 193 performs a process of performing an operation on the data received by the terminal 10 according to a program and outputting the operation result to a memory or the like.

[0076] The notification control unit 194 performs a process of displaying a display image on the display 132, a process of outputting sound to the speaker 142, and a process of generating vibration.

[0077] The storage control unit 195 controls the storage of data in the storage unit 180.

[0078] Explanation will be given for various information stored in the storage unit 180. In a certain aspect, the storage unit 180 stores each information such as user information 181.

[0079] The user information 181 is information of a user who uses the service of the server 20.

[0080] <2 Data Structure> Figure 4 is a diagram showing the data structure of the user database 211. The user database 211 includes an item "user ID", an item "name", an item "email address", an item "attribute", an item "business operator ID", an item "held qualification", an item "registration date", an item "status", an item "field", and an item "evaluation value".

[0081] The item "user ID" is information for identifying each user.

[0082] The item "name" is information indicating the name of the user.

[0083] The item "email address" is information on the email address as the contact information of the user.

[0084] The item "attribute" is information indicating the attribute of the user.

[0085] Specifically, the item "attribute" includes the following as the attribute of the user. · A person who consults on legal matters · An expert who answers in response to legal consultations. For example, a qualified person such as a lawyer The item "business operator ID" is information for identifying the organization to which the user belongs.

[0086] Specifically, the following organizations may exist as the information for identifying the organization to which the user belongs in the item "business operator ID". · Law firm · Business company · Individual (not as a person belonging to an organization, but may consult as an individual) The item "held qualification" is information indicating the qualification held by the user.

[0087] Specifically, the following qualifications may exist as the information for identifying the qualification held by the user in the item "held qualification". · National qualifications such as lawyers, patent attorneys, and certified public accountants · Qualifications recognized by organizations such as business operators and general incorporated associations The item "Registration Date" is information indicating the date and time when the user registered for the service.

[0088] The item "Status" is information indicating the status of the user regarding legal consultations.

[0089] Specifically, the item "Status" includes the following as the status of the user. · Under Consultation: A state where a user who is an expert can receive consultations from the consulters. · Consultation on Hold: A state where a user who is an expert is not receiving consultations. For example, it may be the case that no new consultations are being received because the number of consultations has reached a certain level or more. The server 20 changes the status of the user in the item "Status" according to the user's operation. The item "Field" is information indicating the field in which the user has expertise.

[0090] Specifically, the item "Field" includes information on each field classified as a field of legal consultation. For example, if the server 94 of the information media service provides a service of introducing experts such as lawyers and classifies the fields of experts according to the classification of the consultation content (such as debt restructuring, traffic accidents, divorce, inheritance, etc.), it may include information on each of these classified fields. The server 20 may update the user database 211 according to the user's operation of designating a field, or a user who is an expert may set the field of the user who is an expert by referring to the field of the legal consultation based on the achievements corresponding to the legal consultation.

[0091] The item "Evaluation Value" is information indicating the evaluation value obtained by evaluating the mode of the user's legal consultation.

[0092] Specifically, the item "Evaluation Value" includes the following as the evaluation value of the user. · Evaluation Value as an Expert: The evaluation value evaluated based on the achievements (number of cases, frequency, etc.) of answering consultations from the consulters. · Evaluation value as a counselor: An evaluation value obtained by evaluating based on the achievements of the counselor consulting an expert and being appointed by the expert, and the achievements of the counselor organizing (inputting the consultation content) the consultation content through the legal consultation service provided by the server 20. FIG. 5 is a diagram showing the data structure of the chat consultation history database 212. The chat consultation history database 212 includes an item "consultation thread ID", an item "counselor user ID", an item "post ID", an item "post content", an item "related posts", an item "post date and time", an item "search target database", an item "search result", an item "LLM output", an item "answer", and an item "user evaluation".

[0093] The item "consultation thread ID" is information for identifying the thread of each consultation in which the user consults via chat in the legal consultation service provided by the server 20.

[0094] The item "counselor user ID" is information for identifying the user who conducts the consultation.

[0095] Specifically, the item "counselor user ID" corresponds to the item "user ID" in the user database 211.

[0096] The item "post ID" is information for identifying each post associated with the thread.

[0097] The item "post content" is information indicating the content posted by the user.

[0098] The item "related posts" is information indicating other posts related to the post indicated by the item "post ID" for a plurality of posts managed in the thread.

[0099] Specifically, the item "Related Posts" is information that identifies the most recent post made by the user among a plurality of posts managed in a thread. In this way, each time the user posts a consultation in the thread indicated by the item "Consultation Thread ID", by associating it with the "Post ID" of the most recent post already made and managing it in the chat consultation history database 212, the order of a series of posts in the thread can be managed.

[0100] The item "Post Date and Time" is information indicating the timing when the user made the post.

[0101] Specifically, the item "Post Date and Time" may be the timing when a question was received from the user in free text or the like.

[0102] The item "Database to be Searched" is information that identifies the database to be referred to by search when generating an answer by the server 95 of the large language model service.

[0103] Specifically, the item "Database to be Searched" may include the following information as the database to be referred to. · Information that can be obtained by search on the Internet · Information provided by the server 91 of the case search service · Information provided by the server 92 of the statute search service · Information provided by the server 93 of the book viewing service · Information provided by the server 94 of the information media service · Information provided by the server 95 of the artificial intelligence (large language model) service · Information provided by the server 96 of the business operator · Information provided by the server 97 of the SNS The item "Search Results" is information indicating the search results obtained by searching the database to be referred to based on the content of the user's post.

[0104] Specifically, the item "Search Results" includes the results of the search conducted by the server 20 based on the posted content of the user indicated by the item "Post ID" as follows. · Extract words from the user's posted content and use them as search keys. At this time, based on the legal terms recorded in the legal term database (not shown), for legal terms, they are used as search keys without further decomposition into words. · Analyze the meaning of the user's posted content and vectorize it using methods such as Word2Vec. The documents in the database to be referenced are also vectorized in the same way, and the documents stored in the database are searched based on the proximity of the vector distances. The item "LLM Output" is information indicating the results generated by the server 95 of the large language model service based on the user's post.

[0105] Specifically, the item "LLM Output" includes the results of having the server 95 of the large language model service generate answers as follows. · Results of having the server 95 of the large language model service summarize the results of referring to each database shown in the item "Search Results". · Results of having the server 95 of the large language model service generate answers with backgrounds such as "judgment by users in the legal department" included in the prompt when generating answers. The item "Answer" is information indicating the answer presented to the user.

[0106] Specifically, for the item "Answer", the server 95 of the large language model service may present the user with an answer that includes the results generated as follows and the information generated by the server 20. · An answer including a link that allows reference to the information source of the results summarized by the server 95 of the large language model service. · For the results summarized by the server 95 of the large language model service, an answer with the server 20 adding the original texts of the articles recorded in the statute database, etc., and the cases recorded in the case law database. The item "User Evaluation" is information indicating the user's evaluation of the answer presented by the server 20.

[0107] FIG. 6 is a diagram showing the data structure of the LLM usage history database 213. The LLM usage history database 213 includes an item "usage ID", an item "consultation thread ID", an item "post ID", an item "prompt", an item "LLM output result", an item "usage date and time", and an item "case ID".

[0108] When the server 20 provides a legal consultation service to the user, the server 20 may cause the server 95 of the large language model service to execute a task using the post input by the user. Also, in response to the registration of case data in the case management service provided by the server 20, the server 95 of the large language model service may generate information.

[0109] The item "usage ID" is information for identifying each history of the answers generated by the server 95 of the large language model service.

[0110] The item "consultation thread ID" is information for identifying the thread in which the user consults in the legal consultation service provided by the server 20.

[0111] Specifically, the item "consultation thread ID" corresponds to the item "consultation thread ID" in the chat consultation history database 212.

[0112] The item "post ID" is information for identifying the post of the question input by the user, which is the target for generating an answer by the server 95 of the large language model service.

[0113] Specifically, the item "post ID" corresponds to the item "post ID" in the chat consultation history database 212.

[0114] The item "prompt" is information indicating the content of the prompt applied when transmitting to the server 95 of the large language model service.

[0115] Specifically, the item "prompt" may include the content of the prompt defined as follows and sent to the server 95 of the large language model service. · Based on the content of the question input by the user, search each database in the prompt template defined in the prompt database 214 to be described later, and add the search results (the item "search results" in the chat consultation history database 212) to obtain a prompt to be sent to the server 95 of the large language model service. · Apply the data registered in association with the case in the case management database 216 to the prompt template defined in the prompt database 214 to be described later, and use it as a prompt to be sent to the server 95 of the large language model service. The item "LLM output result" is information indicating the content output by the server 95 of the large language model service by sending a prompt to the server 95 of the large language model service.

[0116] The item "usage date and time" is information on the date and time when the server 95 of the large language model service generated an answer.

[0117] The item "case ID" is information for identifying each case managed in the case management database 216 to be described later.

[0118] Figure 7 is a diagram showing the data structure of the prompt database 214. The prompt database 214 includes an item "prompt ID", an item "operator ID", an item "prompt template", an item "usage permission setting", and an item "availability".

[0119] The item "prompt ID" is information for identifying each prompt.

[0120] The item "operator ID" is information for identifying the organization that uses the prompt.

[0121] Specifically, the item "operator ID" corresponds to the item "operator ID" in the user database 211.

[0122] The item "prompt template" is information indicating the content of the prompt template.

[0123] Specifically, the item "prompt template" may include a prompt template containing the following content. · Summarize the information input by the user in the legal consultation service or the case management database 216. For example, it may be summarized within a certain number of characters. · Arrange the facts related to the consultation content in chronological order in the text of the consultation content of the consulter (for example, the text obtained by speech recognition of the voice data of the consultation content). · Explain the correlation of the people appearing in the consultation content in the text of the consultation content of the consulter in an article. · Extract the next action (next action) from the transcript of the conversation between the consulter and the expert. For example, from the transcript of the conversation, identify keywords such as "summary", "next action", and "next time" as the content related to the next action to be taken, and the content of the speech associated with the keyword may be used as the next action. Also, when there are keywords indicating the next action to be taken, such as "next action", "summary", and "work until next time" in the minutes data, the minutes may be referred to to extract the next action. · Extract what the consulter values as the consultation content and the concerns of the consulter based on the content of the conversation of the consulter and the result of determining the emotion from the voice data. · Extract the concerns of the consulter from the consultation content of the consulter. · Identify the legal arguments from the consultation content of the consulter and the concerns of the consulter. For example, it may be possible to extract words from the consultation content and the concerns of the consulter to identify the arguments. Also, on the server 20, a list of keywords is maintained for each argument, and the words extracted from the consultation content and concerns of the consulter are compared with the list of keywords for each argument to identify the arguments. · Summarize case law, statutes, and descriptions in books regarding legal arguments. For example, identify legal arguments from the concerns of the person seeking advice, search servers 91 for case law search services, servers 92 for statute search services, servers 93 for book viewing services, and other various databases based on these arguments, and summarize the search results. It may also extract the basis articles corresponding to the arguments. · Search for case law. Point out changes in the interpretation of case law corresponding to the argument. For example, when amending and indicating the original judgment, display the amended parts in the judgment text of the higher court with respect to the original judgment. · Compare the past interpretation and the latest interpretation of the interpretation of case law and point out the differences. For example, search for specific cases of the argument and contrast the interpretation recorded in books published in the past with the latest interpretation. · Based on the information stored in the case management database 216, create legal-related documents such as contracts, term sheets, warning letters, response letters, pleadings, pleadings, and preparatory documents. For example, it may be possible to create a pleading by referring to the information of the parties. · For legal documents such as pleadings and preparatory documents, output the counterarguments of the assumed counterpart to the claims related to the argument. For example, search for past case law regarding the argument, and based on these past case law, generate the assumed counterarguments on the server 95 of the large language model service. · Output the citation relationship of the terms of the contract according to a predetermined format. For example, it may be possible to output the terms of the citation source and the citation destination of the contract in pairs. · Generate a list of the materials and evidence registered in the case management database 216. · Generate a list of the evidence mentioned in the data such as pleadings and preparatory documents registered in the case management database 216. · A form for generating an answer on the server 95 of the large language model service. For example, there are those that specify the items to be included in the answer, and there may be a prompt that instructs to generate an answer in accordance with the specified items. Such items may include legal arguments, a list of interested parties, generating answers separately for each interested party, the advantages and disadvantages of the interested parties, etc. · An instruction to generate a summary for data stored in the operator's server 96, which is not publicly available to third parties and is referenced by users belonging to the organization The item "Usage Permission Setting" is information indicating the range of users who have the right to generate answers on the server 95 of the large language model service using prompts.

[0124] Specifically, the item "Usage Permission Setting" may include information on permissions set as follows. · When generating a summary for data that is not publicly available to third parties, such as data stored in the operator's server 96, users who do not belong to the organization have no permission, and permissions are set for users who belong to the organization · Set permissions according to the user's department (for example, generate a summary for data that can be referenced by specific departments such as business planning) · Set permissions according to the user's position (for example, generate a summary for data that can be referenced by users in specific positions such as management) The item "Availability" is information indicating whether to use the prompt template as a target for use or not.

[0125] Specifically, the item "Availability" includes information on whether to permit or not permit the use of the prompt template. For example, the server 95 of the large language model service may update the prompt template for generating better answers and stop the use of the previous prompt template.

[0126] Figure 8 is a diagram showing the data structure of the expert consultation history database 215. The expert consultation history database 215 includes an item "Expert Consultation ID", an item "Chat Consultation ID", an item "Consultation Content", an item "Field", an item "Expert ID", an item "Consultation Date and Time", and an item "Assignment Result".

[0127] The item "Expert Consultation ID" is information for identifying each request when a consulter requests consultation from an expert.

[0128] Specifically, the item "Expert Consultation ID" corresponds to the item "Consultation with Expert" in the chat consultation history database 212.

[0129] The item "Chat Consultation ID" is information that identifies each consultation conducted by the consulter in a chat in the legal consultation service provided by the server 20 when the consulter requests a consultation from an expert.

[0130] Specifically, the item "Chat Consultation ID" corresponds to the item "Chat Consultation ID" in the chat consultation history database 212.

[0131] The item "Consultation Content" is information indicating the content of the consultation requested by the consulter from an expert.

[0132] Specifically, the item "Consultation Content" includes information on the content of the consultation request entered by the consulter at the stage when the consulter performs an operation to request a consultation from an expert. The server 20 may obtain the content entered by the consulter into the chat system by referring to the chat consultation history database 212, and have the server 95 of the large language model service summarize at least one of the consulter's consultation content or the answer generated by the server 95 of the large language model service for the question, and the summarized result may be used by the consulter as information on the content of the request when the consulter requests a consultation from an expert. Thereby, it can be made even easier for the consulter to input the content to be requested from an expert.

[0133] The item "Field" is information indicating the field of the consultation set for the consulter's request for a consultation from an expert.

[0134] Specifically, the item "Field" includes information on the field set by the server 20 or the user (consulter or expert).

[0135] In addition, for example, the server 20 may identify the field of consultation by extracting legal terms included in the content of the user's question in the chat consultation history database 212 and the answer generated by the server 95 of the large language model service.

[0136] The item "Expert ID" is information for identifying the expert who responded to the consultation request from the consultor.

[0137] Specifically, the item "Expert ID" corresponds to the item "User ID" in the user database 211.

[0138] The item "Consultation Date and Time" is information on the date and time when the consultation request from the consultor to the expert was registered.

[0139] The item "Acceptance Result" is information indicating whether the expert has accepted the consultation request from the consultor.

[0140] Specifically, the item "Acceptance Result" includes the following information. · Accepted: The expert has reached the acceptance of the consultation request from the consultor (for example, the expert has registered that they have reached the acceptance). · Not Accepted: The expert responded to the consultation request from the consultor but did not reach the acceptance (timed out after a certain period from the consultation date and time, and the expert registered that they did not reach the acceptance). Figure 9 is a diagram showing the data structure of the case management database 216. The case management database 216 includes an item "Case ID", an item "Consultor User ID", an item "Expert User ID", an item "Consultation Audio Data", an item "Consultation Content Memo", an item "Consultation Date and Time", an item "AI Summary of Consultation Content", an item "AI Argument Extraction", an item "AI Research Result", an item "Document Data", and an item "Legal Document Data".

[0141] The item "Case ID" is information for identifying each case.

[0142] The item "Consultant User ID" is information that identifies the user who is the consultant.

[0143] Specifically, the item "Consultant User ID" corresponds to the item "User ID" in the user database 211.

[0144] The item "Expert User ID" is information that identifies the user who is the expert.

[0145] Specifically, the item "Expert User ID" corresponds to the item "User ID" in the user database 211.

[0146] The item "Consultation Voice Data" is the voice data of the conversation when the consultant consults an expert.

[0147] Specifically, the item "Consultation Voice Data" may include the following. · The voice data itself · The text transcription data obtained by performing character recognition processing on the voice data · The data of the speaker's emotion obtained by performing speaker and emotion determination processing on the voice data The item "Consultation Content Memo" is the memo information input by the expert or the consultant.

[0148] The item "Consultation Date and Time" is the information on the time when the consultation from the consultant to the expert was made.

[0149] Specifically, the item "Consultation Date and Time" is The item "Consultation Content AI Summary" is the information on the result summarized by the server 95 of the large language model service for the data of the consultation content of the consultant.

[0150] Specifically, the item "Consultation Content AI Summary" includes the result summarized by the server 95 of the large language model service based on the text transcription data of the voice data related to the conversation between the consultant and the expert.

[0151] The item "AI Argument Extraction" is information on the arguments identified by the server 95 of the large language model service for the data of the consultation content of the consulter.

[0152] The item "AI Research Result" is information on the result of summarizing information such as case laws, statutes, interpretations, and commentaries regarding the arguments by the server 95 of the large language model service.

[0153] Specifically, the item "AI Research Result" includes the result of searching various databases such as the server 91 of the case search service, the server 92 of the statute search service, and the server 93 of the book browsing service for the arguments and having the search results summarized by the server 95 of the large language model service.

[0154] The item "Document Data" is information such as the information provided by the consulter regarding the case and the information collected by the experts.

[0155] Specifically, the item "Document Data" includes information such as documents not disclosed to the other party, documents generally published, and evidence submitted in legal procedures.

[0156] The item "Legal Document Data" is data of the documents submitted in legal procedures.

[0157] Specifically, the item "Legal Document Data" includes the following types of document data. · Contract · Term Sheet · Warning Letter · Reply Letter (responding to the warning letter) · Complaint · Answer · Preparation Document <3 Operations (First Embodiment)> Figure 10 is a flowchart showing the process flow in which the user requests consultation from an expert through legal consultation in chat and the expert accepts the appointment.

[0158] In step S1021, the server 20 presents an operation screen for receiving input of questions about legal consultations to the user who is the consulter.

[0159] In step S1011, the terminal 10 displays an answer generated by the server 95 of the large language model service for the question of legal consultation received from the user. The terminal 10 receives from the user an operation of requesting consultation with an expert about the legal consultation related to the question on the operation screen.

[0160] The server 20 is configured to store in the user database 211 information of experts such as lawyers who can answer legal consultations. The server 20 refers to the chat consultation history database 212 and extracts candidates for experts corresponding to the consultation according to the request based on the content of the consultation related to the user's request, the answer generated by the server 95 of the large language model service, and the information of the experts. For example, the server 20 may identify the field related to the consultation content from the terms included in the consultation content and the content of the generated answer, and extract lawyers specializing in the field from the user database 211.

[0161] The terminal 10 displays the experts extracted by the server 20 as candidates for requesting consultation. The terminal 10 presents to the user the candidates for the extracted experts together with the information of the experts (for example, name, field of expertise, etc.).

[0162] The terminal 10 may receive input of conditions for selecting an expert from the user. For example, there may be a case where it is desired to further narrow down the candidates for experts extracted by the server 20 or search for experts different from the extracted experts. The server 20 may extract candidates for experts according to the conditions received from the user and the candidates for experts.

[0163] The terminal 10 may also accept an operation from the user to publicly disclose questions about the user's legal consultation so that a third party can view them. The server 20 may, in response to the disclosure operation, make the questions about the legal consultation viewable by a third party. When making the questions and answers viewable by a third party, the server 20 may generate an answer by instructing the server 95 of the large language model service not to include the user's personal information in the generated answer.

[0164] In step S1023, based on the chat consultation history database 212, the server 20 outputs to the expert user that there is a consultation request from the consulter and the content based on the answer generated by the server 95 of the large language model service for the consultation.

[0165] Here, the server 20 may also generate a prompt instructing to generate an answer for the question of the user who is the consulter in accordance with the specified items, and cause the server 95 of the large language model service to generate an answer. For example, the server 20 may generate a prompt instructing to generate an answer by arranging as follows for the specified items. · Items of interested parties · Items of time information · Items of arguments · Items of monetary value related to the legal consultation (such as how much the damages are) · Items of relevant laws and regulations, precedents The server 20 sends a prompt instructing to generate an answer in accordance with the specified items as described above to the server 95 of the large language model service, receives the generated answer, and updates the chat consultation history database 212.

[0166] Server 20 may output content based on the answers generated by experts in accordance with the specified items. For example, Server 20 manages the case management database 216 used by experts, and creates a prompt to generate answers in accordance with the specified items as described above so as to match the items of the records that manage each case in the case management database 216. Thereby, Server 20 may update the case records in the case management database 216 based on the results generated by the server 95 of the large language model service.

[0167] Server 20 evaluates the user's consultation based on the questions input by the user and the answers generated by the server 95 of the large language model service, which are held in the chat consultation history database 212, and updates the item "acceptance possibility" of the chat consultation history database 212 based on the evaluation result.

[0168] Server 20 may refer to the chat consultation history database 212 and output, to the expert, the score for evaluating the user's consultation regarding the request of the user who is the consulter.

[0169] Server 20 may refer to the expert consultation history database 215 and evaluate the possibility of acceptance according to the record of whether the legal consultation of the consulter has led to acceptance by the expert.

[0170] Server 20 may refer to the chat consultation history database 212 and evaluate the possibility of acceptance based on at least any one of the data volume of the questions input by the user regarding the legal consultation, whether the large language model has succeeded in answering, and whether the answers generated by the large language model have a data volume above a certain level.

[0171] Server 20 may refer to the chat consultation history database 212 and preferentially output to the expert the legal consultations regarding requests that are evaluated to have a possibility of acceptance above a certain level.

[0172] In step S1025, when the server 20 receives an operation from an expert in response to a request from a counselor, it updates the expert consultation history database 215 and notifies the user who is the counselor.

[0173] In step S1013, the terminal 10 presents to the user that the expert has responded to the request.

[0174] As described above, an example has been explained in which an expert receives a consultation from a counselor by matching with a counselor who uses a legal consultation service and registers the information of the case in the case management database 216. However, this is not the only case, and there may be a case where an expert receives an inquiry from a counselor and manages the case in the case management database 216.

[0175] FIG. 11 is a diagram showing the flow of processing for summarizing the consultation content from a counselor by a large language model and associating it with the record of the case.

[0176] In step S1121, the case management module 2048 of the server 20 outputs an operation screen for managing the case to the terminal 10.

[0177] In step S1111, the terminal 10 receives an operation of uploading consultation data indicating the consultation content from the counselor on the operation screen. The consultation data is, for example, voice data, recording data of a meeting, a memo input in an interview between the counselor and the expert, etc.

[0178] In step S1123, the case management module 2048 of the server 20 receives the consultation data from the terminal 10 and registers it in the record of the case in the case management database 216.

[0179] The case management module 2048 receives, as consultation data, voice data indicating the content of the consultation. When the case management module 2048 receives the voice data, it performs character recognition processing to obtain the transcribed data obtained by transcribing the voice data. The case management module 2048 registers the transcribed data as case information by associating it with the counselor and storing it in the records of the case management database 216.

[0180] The case management module 2048 performs speaker determination and sentiment determination processing on the voice data, and further stores in the case management database 216 at least any one of the information on the result of determining the speaker in the transcribed data, the information on the result of determining the sentiment, and the information on the timing when the consultation from the counselor occurred.

[0181] In step S1125, the case management module 2048 of the server 20 transmits a prompt for summarizing the consultation content indicated in the consultation data along a predetermined item to the server 95 of the large language model service. Here, the predetermined item may correspond to each subfield indicated in the case management database 216.

[0182] The case management module 2048 may obtain the result of summarizing the consultation content by giving a prompt to the server 95 of the large language model service so as to summarize the consultation content indicated in the consultation data along a predetermined item.

[0183] The case management module 2048 may obtain the result of summarizing in time series by giving a prompt to the server 95 of the large language model service to sort and organize the consultation content included in the consultation data in time series as a prompt.

[0184] For example, the case management module 2048 may cause the server 95 of the large language model service to generate a text in which the consultation content is sorted in time series by giving the following prompt to the server 95 of the large language model service. ·Identify the date of year, month, and day included in the transcription of the consultation content, and extract the consultation content (facts, etc.) associated with the date of year, month, and day (for example, statements such as "There was ~~~ around ~~ year"). ·Based on the expression of the past tense of the verb, extract the facts that occurred in the past (for example, from the statement content such as "Loaned money", determine that it is an expression in the past tense of "loaned", and extract it as a fact that occurred in the past). ·When there is no word indicating the time period for a fact, treat it as a fact of the time period detected most recently. ·Extract the relative time period by detecting the demonstrative pronoun and the word indicating the time period (for example, if there is a statement such as "two months before that", detect that it is a time period that is relatively before or after compared to the time period represented by the demonstrative pronoun (for example, the most recently detected time period). For example, if it is "two months before that", it indicates a time period further before than the time period indicated by the demonstrative pronoun). ·When there is no demonstrative pronoun, use the fact based on the time point when the consultation occurred as the reference. In step S1127, the case management module 2048 of the server 20 obtains the result of summarizing the consultation content from the server 95 of the large language model service, and registers it in the record of the case for the counselor in the case management database 216. The case management module 2048 associates the consultation content summarized along with the predetermined items by the server 95 of the large language model service with the counselor, and stores it in the case management database 216 so that the expert responding to the consultation from the counselor can refer to it.

[0185] When the case management module 2048 obtains the summarized result sorted in chronological order from the server 95 of the large language model service, it stores it in the case management database 216 in association with the counselor.

[0186] In the case management database 216, each consultation from the counselor is managed as a case including predetermined management items. The case management module 2048 may associate and store each item of the consultation content summarized along with the predetermined items with each management item managed as a case in the case management database 216.

[0187] In the case management database 216, at least one of the consultation content from the consulter, the next action which is the action to be taken next, and the concerns of the consulter is managed as a case as a predetermined management item. The case management module 2048 may obtain the summarized result by giving a prompt to the server 95 of the large language model service so as to summarize at least one included in the predetermined management item as a predetermined item.

[0188] The case management module 2048 may extract the concerns of the consulter based on the voice data of the consultation content of the consulter and the information of the result of determining the emotion, and store the information of the concerns of the consulter as the information of the case in the case management database 216. For example, based on the result of determining the emotion, in the following cases, it may be extracted as the concerns of the consulter. · The text corresponding to the voice of the consultation content of the consulter while the emotion is determined to be "anger" or "anxiety" · The text corresponding to the voice of the consultation content of the consulter before and after the evaluation value of the evaluated emotion changes to a certain level or more (the emotion fluctuates) · The text corresponding to the voice in a state where the volume of the voice data of the consulter is a certain level or more (it may also include the front and back thereof) (such as speaking in a loud voice) · When the speech-to-text of the voice data of the consulter includes a statement indicating something to be concerned about. For example, as statements, there may be "worried", "want to do something about it" (a statement with the intention of solving a problem), "hate", "angry" (a statement showing a negative emotional reaction), "don't want to be ~~" (a statement with the intention of avoiding falling into a specific behavior or a specific state), "is it okay to do ~~~" (a statement with the intention of confirming whether it is okay to shift to a specific behavior or a specific state. For example, "is it okay to negotiate directly with the other party"), etc. · Among the keywords appearing in the consulter's speech, those with the number of appearances being a certain level or more or with a high frequency (because of repeated speech, there may be concerns about the matters corresponding to the keywords) The case management module 2048 may cause the server 95 of the large language model service to extract the information on the concerns of the counselor by providing the server 95 of the large language model service with a prompt, the voice data of the counselor, the speech-to-text of the voice data, the result of sentiment determination, etc., so as to extract the concerns of the counselor as described above.

[0189] In step S1129, the case management module 2048 of the server 20 outputs the information on the record of the case to the terminal 10.

[0190] In step S1113, the terminal 10 displays the information on the record of the case.

[0191] FIG. 12 is a diagram showing the process flow of generating document data by receiving an operation of decomposing and rearranging the consultation content from the counselor.

[0192] In step S1221, the case management module 2048 of the server 20 outputs an operation screen to the terminal 10, which decomposes the consultation content of the counselor and displays each as an object that the user can rearrange.

[0193] In step S1211, the terminal 10 displays the information on the record of the case. The terminal 10 displays the operation screen and receives an operation from the user to rearrange each object corresponding to the consultation content.

[0194] In step S1223, the case management module 2048 of the server 20 provides the server 95 of the large language model service with a prompt to generate a text in chronological order for the consultation content shown in each object based on the order of rearrangement according to the operation of rearranging the objects. Thereby, the case management module 2048 obtains a summarized result sorted in chronological order for the consultation content from the counselor and registers it in the case management database 216. The case management module 2048 outputs the result summarized in chronological order to the terminal 10.

[0195] In step S1213, the terminal 10 displays the results summarized by the server 95 of the large language model service, which are sorted in chronological order.

[0196] FIG. 13 is a diagram showing the flow of processing for expressing a person correlation diagram in text.

[0197] In step S1311, the terminal 10 displays an operation screen and accepts an upload of an image of a person correlation diagram related to the consultation content from the user.

[0198] In step S1323, the case management module 2048 of the server 20 transmits a prompt for generating a text showing the correlation between persons to the server 95 of the large language model service based on the received image of the person correlation diagram.

[0199] Here, the person correlation diagram is a diagram showing the relationships between the characters who appear.

[0200] Specifically, the case management module 2048 causes the server 95 of the large language model service to generate a text showing the relationship between persons by giving, as a prompt, a rule for interpreting the image of the person correlation diagram to the server 95 of the large language model service as follows.

[0201] For example, the prompt includes the following instruction content. · Figures such as circles and rectangles represent the characters who appear. If the figure contains a person's name or a person's image, the figure represents the character who appears · If text of a person's name is placed near the figure, the figure represents the character who appears and the text represents the name of the character who appears · An arrow connecting figures such as circles and rectangles represents the subject who performed the action and the recipient who received the action. · If text is arranged in association with the arrow, the text indicates the content of the action. · If the line connecting figures such as circles and rectangles is an equal sign, it simply indicates the relationship between persons. · If text is arranged in association with the equal sign line, the text indicates the content of the relationship between people. · Based on the above rules, please describe the given figure in words. In step S1325, the case management module 2048 of the server 20 obtains the result of generating a text showing the correlation of people from the server 95 of the large language model service, and registers it in the record of the case for the counselor in the case management database 216.

[0202] In step S1327, the case management module 2048 of the server 20 outputs a text showing the correlation of people to the terminal 10 as information in the record of the case in the case management database 216.

[0203] In step S1313, the terminal 10 displays a text showing the correlation of people as information in the record of the case.

[0204] Figure 14 is a diagram showing the process of presenting questions to the user to enhance the resolution of information about people.

[0205] In step S1411, the terminal 10 accepts the upload of voice data of the consultation content.

[0206] In step S1423, the case management module 2048 of the server 20 obtains the speech-to-text conversion of the voice data of the consultation content and registers it in the record of the case in the case management database 216.

[0207] In step S1425, when the case management module 2048 of the server 20 identifies a word indicating a person in the speech-to-text data, it presents a question asking about the relationship between the identified person and other characters involved in the consultation content to the user of the counselor or expert, and accepts the input of the answer.

[0208] For example, when the case management module 2048 identifies a word consisting of a proper noun and an honorific (such as "san" which is an honorific added to a name) in the transcribed data, it may also identify the word as indicating a person. Further, the case management module 2048 may identify the word indicating a person by collating with a dictionary indicating personal names in the transcribed data.

[0209] In step S1413, the terminal 10 receives from the user an answer to a question asking about the relationship between people.

[0210] In step S1427, the case management module 2048 of the server 20 registers the answer regarding the relationship between people in the record of the case in the case management database 216. The case management module 2048 outputs the information of the record including the sentence indicating the relationship between people to the terminal 10.

[0211] In step S1415, the terminal 10 displays the information of the record of the case.

[0212] <4 Screen example (First embodiment)> FIG. 15 is an example of an operation screen for a user who is a counselor to request counseling from an expert.

[0213] The operation screen 1200 is a screen for receiving legal counseling in a chat format from a user who is a counselor.

[0214] The account display area 1202 is an area for displaying the information of the account of the user who is a counselor.

[0215] The chat display area 1204 is an area for displaying the questions input in the legal counseling and the answers generated by the server 95 of the large language model service.

[0216] The navigation display area 1212 is an area for displaying information to assist the user in counseling in a chat format regarding legal counseling.

[0217] In the illustrated example, the navigation display area 1212 presents to the user the fact that an answer has been generated for the user (the "AI answer display area 1222") and the content of the operation for making a request to an expert.

[0218] The chat input reception unit 1214 is an area that receives input of a question by accepting input of text from the user.

[0219] As shown in the figure, the chat input reception unit 1214 may display information suggesting the content to be input in order to prompt the user to input a question sentence. This corresponds to step S1011 and the like in FIG. 10.

[0220] The AI answer display area 1222 is an area that displays the answer generated by the server 95 of the large language model service for the question sentence input by the user.

[0221] The request display area 1224 is an area that displays a screen for requesting an expert regarding legal consultation.

[0222] The request display area 1224 corresponds to step S1011 and the like in FIG. 10.

[0223] The candidate request destination display area 1226 is an area that displays candidates for the request destination lawyer.

[0224] The candidate request destination display area 1226 corresponds to step S1011 and the like in FIG. 10. In the illustrated example, the candidate request destination display area 1226 is accepting an operation from the user to select a lawyer to request.

[0225] The condition specification reception unit 1228 is an operation member that receives specification of conditions for extracting candidate lawyers to whom a consultation is to be requested.

[0226] The condition specification reception unit 1228 corresponds to step S1011 and the like in FIG. 10.

[0227] The request operation reception unit 1230 is an operation member that receives an operation to request advice from a lawyer.

[0228] The request operation reception unit 1230 corresponds to step S1011 in FIG. 10 and the like.

[0229] FIG. 16 is an example of an operation screen on which an expert receives a consultation request from a consulter and performs an operation.

[0230] The operation screen 1232 is a screen that displays the consultation content from the consulter to the user who is an expert and receives an operation in response to the request.

[0231] The account display area 1234 is an area that displays the information of the account of the user who is an expert.

[0232] The request content display area 1236 is an area that displays the content of the consultation from the consulter.

[0233] The detailed display area 1238 is an area that displays the details of the consultation from the consulter.

[0234] As shown in the figure, the detailed display area 1238 displays the consultation from the consulter in a selectable manner for the expert. It corresponds to step S1023 in FIG. 10 and the like. In the detailed display area 1238, as shown in the figure, the server 20 refers to the chat consultation history database 212 and displays values obtained by scoring the consultation and the like.

[0235] The contact operation reception unit 1240 is an operation member that receives an operation to contact the consulter in response to a consultation request from the consulter.

[0236] The contact operation reception unit 1240 corresponds to step S1025 in FIG. 10 and the like.

[0237] The corresponding case display area 1242 is an area that displays the cases that the expert has received a request from the consulter and is corresponding to (undertaking).

[0238] The detailed display area 1244 is an area for displaying the details of the cases that the experts are handling.

[0239] The search operation reception unit 1246 is an operation member that receives an operation to search for the cases for which requests have been received.

[0240] Figure 17 is an example of a screen for displaying the records of cases.

[0241] The operation screen 1300 is a screen that receives operations for managing cases.

[0242] The account display area 1302 is an area for displaying the information of the user's account.

[0243] In the illustrated example, the account display area 1302 displays information for identifying the logged-in user.

[0244] The new case registration area 1304 is an area that receives an operation to register a new case.

[0245] The new case registration area 1304 receives the registration of data for each sub-field of the case management database 216 and updates the case management database 216.

[0246] The counselor designation unit 1306 is an operation member that receives the designation of information about the counselor related to the case.

[0247] When there is a master in which the information of the counselor has already been registered, the counselor designation unit 1306 searches the master according to the input of the user's text or the like. In the example of Figure 18, the information of the counselor is registered in the counselor designation unit 1306.

[0248] The interview data designation unit 1308 is an operation member that receives the designation of data during the interview between the counselor and the expert.

[0249] The interview data specifying unit 1308 accepts uploads of voice data, video data, etc. as data during the interview. Additionally, although not shown in the figure, it also accepts registration of notes related to the case. In the example of FIG. 18, it shows that the upload of video data has been accepted.

[0250] The case list display area 1310 is an area for displaying a list of cases.

[0251] The case list display area 1310 reads and displays records from the case management database 216 according to the user's viewing authority. Also, the server 20 accepts an operation to search for cases by the information of the counselor, etc., and displays the search results in the list display area 1312, etc. In the example of FIG. 18, it shows that a new record has been added to the case management database 216 due to accepting the registration of a new case in the new case registration area 1304. Also, in the example of FIG. 18, it shows that, by using the information registered in the record of the case after accepting the registration of a new case, a prompt is given to the server 95 of the large language model service, and the response result is registered in the record of the case.

[0252] The list display area 1312 is an area for displaying a list of records of cases.

[0253] The list display area 1312 accepts an operation from the user to specify each record, and displays detailed information of the specified record. In the example of FIG. 18, in the record of the newly registered case, the speech-to-text conversion of the voice data is displayed in the speech-to-text viewing operation unit 1314 (corresponding to step S1123 in FIG. 11). Also, in the example of FIG. 18, the summarization of the consultation content by the server 95 of the large language model service is displayed in the summary viewing operation unit 1316 (corresponding to steps S1125 and S1127 in FIG. 11).

[0254] FIG. 18 is an example of a screen showing the result of generating a summary of the consultation content based on the voice data of the consultation content in a case.

[0255] The text conversion viewing operation unit 1314 is an operation member that accepts an operation to view the text conversion data obtained by converting the voice data of the consultation content into text.

[0256] The summary viewing operation unit 1316 is an operation member that accepts an operation to view the result of summarizing the consultation content.

[0257] The person correlation display area 1318 is an area for displaying details of information about the persons related to the consultation content.

[0258] In the illustrated example, the person correlation display area 1318 displays the person names identified by the server 95 of the large language model service based on the consultation content. Also, in the illustrated example, the user is being asked about the relationship between the persons, and the answers are being received. This corresponds to step S1425 in FIG. 14 and the like.

[0259] The extracted person display area 1320 is an area for displaying the person names identified from the consultation content.

[0260] In the illustrated example, the extracted person display area 1320 displays that a plurality of persons have been identified.

[0261] The correlation specifying unit 1322 is an operation member that accepts the specification of an answer to a question asking about the relationship between persons.

[0262] In the illustrated example, the correlation specifying unit 1322 accepts an input from the user regarding the relationship between the persons who appear.

[0263] FIG. 19 is an example of an operation screen that accepts an operation to organize the consultation content.

[0264] The summary result display area 1324 is an area for displaying the result of summarizing the consultation content.

[0265] In the example of FIG. 19, the summary result display area 1324 decomposes the result of summarizing the consultation content, for example, by sentence unit, and displays each as an object. In the example of FIG. 19, the user is prompted to perform an operation of rearranging the objects along the time series, and it is displayed that a sentence is output along the rearranged order (corresponding to step S1221 in FIG. 12, etc.).

[0266] In the example of FIG. 20, the summary result display area 1324 supports the creation of sentences in the person correlation diagram.

[0267] The first object 1326 is the result of summarizing the consultation content.

[0268] The first object 1326 accepts an operation from the user to perform an operation of rearranging the positions with the second object 1328 and the third object 1330.

[0269] The second object 1328 is the result of summarizing the consultation content.

[0270] The third object 1330 is the result of summarizing the consultation content.

[0271] The generation operation unit 1332 is an operation member that accepts an operation of generating a sentence based on the object with the specified order.

[0272] In the illustrated example, the generation operation unit 1332 accepts an operation from the user to cause the server 95 of the large language model service to generate a sentence that arranges the facts related to the consultation content in time series, assuming that there are facts of the first object 1326, the second object 1328, and the third object 1330 in time series as the order.

[0273] FIG. 20 is an example of an operation screen that accepts an operation of organizing the correlation of the characters appearing.

[0274] The correlation image specifying unit 1334 is an operation member that accepts the specification of an image to be uploaded.

[0275] The correlation image specifying unit 1334 accepts the upload of the image of the person correlation diagram (corresponding to steps S1311, S1323, etc. in FIG. 13). As shown in the figure, an explanatory text for explaining the method of creating the person correlation diagram may be displayed near the correlation image specifying unit 1334. For example, as described in the prompt explanation in step S1323 above, a text explaining the roles of figures, texts, etc. used in the person correlation diagram (for example, "The characters are represented by figures such as rectangles or circles", "The names of the characters are described in association with the figures", "The relationship between the characters is represented by an arrow and the text associated with the arrow") is displayed near the correlation image specifying unit 1334 (in the illustrated example, the summary result display area 1324). Thereby, the user can easily create a person correlation diagram and can easily obtain the text indicating the correlation between the people.

[0276] In addition to uploading images, the correlation image specifying unit 1334 may also accept the upload of files of presentation applications and files created by document tools. For example, by acquiring these files and processing them as images, etc., a text indicating the correlation between people can be generated based on the figures, etc. included in these files.

[0277] The generation operation unit 1336 is an operation member that accepts an operation to cause the server 95 of the large language model service to generate a text of person correlation based on the image uploaded by the correlation image specifying unit 1334.

[0278] According to the user's operation, the generation operation unit 1336 causes the server 20 to send a prompt for generating a text indicating the correlation of people for the image accepted for upload by the correlation image specifying unit 1334 to the server 95 of the large language model service. Thereby, the server 20 can obtain the text indicating the correlation of the person from the server 95 of the large language model service.

[0279] The correlation relationship display area 1338 is an area that displays a sentence indicating the correlation of a person, which is generated according to an operation on the generation operation unit 1336.

[0280] The correlation relationship display area 1338 corresponds to steps S1327, S1313, etc. in FIG. 13.

[0281] <5 Operations (Second Embodiment)> The second embodiment will be described. After an expert hears the consultation content from the consulter, the consultation content is sorted out. When creating legal documents such as contracts, warning letters, pleadings, and answers, it is necessary to consider what claims to make. However, it is sometimes difficult to estimate how much the workload of the investigation for this purpose will be, which places a burden on the expert side. Also, the investigation work may be postponed. Therefore, in the second embodiment, an example in which the server 20 supports sorting out the arguments and conducting investigations corresponding to the arguments based on the consultation content will be mainly described. FIG. 21 is a diagram showing the flow of a process of specifying the arguments regarding the consultation content of the consulter by a large language model and associating and holding them with the record of the case.

[0282] In step S2121, the case management module 2048 of the server 20 gives a prompt for extracting the arguments corresponding to the consultation content to the server 95 of the large language model service based on the information of the consultation content associated with the record of the case in the case management database 216. As a result, the case management module 2048 acquires the information on the legal arguments regarding the consultation content from the server 95 of the large language model service and updates the case management database 216.

[0283] The case management module 2048 identifies the concerns of the counselor regarding the consultation content based on the information of the consultation content in the case management database 216. For example, the case management module 2048 provides a prompt to the server 95 of the large language model service to point out the concerns of the counselor regarding the consultation content. Thereby, the case management module 2048 may also obtain information on the legal arguments regarding the consultation content by identifying the concerns of the counselor regarding the consultation content and providing a prompt to the server 95 of the large language model service to point out the arguments corresponding to the identified concerns.

[0284] In step S2123, the case management module 2048 of the server 20 searches multiple information sources related to law, such as the statute database (server 92 of the statute search service), the case database (server 91 of the case search service), the legal book database (server 93 of the book viewing service), and other information sources related to law (server 94 of the information media service, server 97 of the SNS, the Internet, etc.) for the arguments associated with the records of the cases in the case management database 216 using the arguments obtained by the process in step S2121.

[0285] In step S2125, the case management module 2048 of the server 20 causes the server 95 of the large language model service to summarize the results of searching multiple information sources, and stores the summarized results in association with the records of the cases related to the arguments in the case management database 216.

[0286] For example, the case management module 2048 may search the server 91 of the case search service for the arguments, and may detect that the appellate court has amended the judgment of the lower court in the searched cases. The case management module 2048 may also present the amendment content of the judgment amended by the appellate court related to the detection to the user.

[0287] In addition, the case management module 2048 searches for information stored in multiple information sources for the first information source that becomes past information in chronological order and the second information source that is future in chronological order than the first information source (for example, referring to the date when various data stored in the database was updated), and gives a prompt to the server 95 of the large language model service to point out the differences by comparing the result of searching the first information source with the result of searching the second information source, so that the server 95 of the large language model service outputs the differences and stores them in the case management database 216. The case management module 2048 may also present the output result of the differences to the user.

[0288] FIG. 22 is a diagram showing the flow of a process for generating a list of materials and evidence registered in a case record.

[0289] In step S2211, the terminal 10 receives an operation from the user to register materials and evidence related to the consultation content on the operation screen.

[0290] In step S2221, the case management module 2048 of the server 20 receives the registration of materials and evidence from the user and registers the information of the materials and evidence in the case record of the case management database 216. In the case management database 216, materials, evidence, preparatory documents, and other written data used in litigation procedures related to the consultation content of the case are managed in association with the case record.

[0291] In step S2223, the case management module 2048 of the server 20 sends a prompt to list these various data to the server 95 of the large language model service based on the data such as materials, evidence, and preparatory documents in the case management database 216. In this way, the case management module 2048 causes the large language model to generate information indicating a list of these based on various written data registered in the case management database 216.

[0292] In step S2225, the case management module 2048 of the server 20 obtains information indicating a list from the server 95 of the large language model service and registers it in the record of the case for the counselor in the case management database 216.

[0293] In step S2227, the case management module 2048 of the server 20 extracts the evidence described in the preparatory document based on the data of the preparatory document. For example, the case management module 2048 identifies the locations (such as the text indicating the evidence) in the data of the preparatory document that mention the evidence.

[0294] In step S2229, the case management module 2048 of the server 20 refers to the case management database 216 and outputs the list information and the result of extracting the evidence described in the preparatory document to the terminal 10.

[0295] In step S2213, the terminal 10 displays the list information of materials, evidence, etc. and the extraction result of the evidence described in the preparatory document.

[0296] <6 Screen example (Second embodiment)> FIG. 23 is an example of an operation screen for displaying information extracted by a large language model in the record of a case.

[0297] The case list display area 1340 is an area for displaying a list of cases.

[0298] The list display area 1342 is an area for displaying a list of case records.

[0299] In the example of FIG. 23, the argument display area 1342 displays, in the argument browsing operation unit 1344, the arguments identified by the server 95 of the large language model service (corresponding to step S2121 in FIG. 21). Also, in the example of FIG. 23, the research browsing operation unit 1346 displays the result of searching for a plurality of information sources corresponding to the arguments by the server 95 of the large language model service and summarizing the results (corresponding to steps S2123 and S2125 in FIG. 21).

[0300] The argument browsing operation unit 1344 is an operation member that accepts an operation to confirm the arguments identified by the server 95 of the large language model service.

[0301] The argument browsing operation unit 1344 displays the arguments corresponding to the consultation content according to the user's operation.

[0302] The research browsing operation unit 1346 is an operation member that accepts an operation to search for a plurality of information sources about an argument and confirm the summary of the search results.

[0303] In the example shown in the figure, as a result of searching for case law, the server 95 of the large language model service outputs a notice that there has been a change in the interpretation of the case law. In response to this result, the research browsing operation unit 1346 displays a notice that the judgment has been revised.

[0304] FIG. 24 shows an example of an operation screen for displaying a list of materials and evidence in the record of a case.

[0305] The evidence list display area 1348 is an area for displaying a list of evidence associated with the record of the case specified by the user.

[0306] The evidence list display area 1348 corresponds to steps S2223, S2225, etc. in FIG. 22.

[0307] The evidence acquisition unit 1350 is an operation member that accepts an operation to acquire evidence information in association with a case.

[0308] The evidence acquisition unit 1350 accepts the registration of various data serving as evidence, such as documents, images, and voices, corresponding to steps S2211 and S2221 in FIG. 22.

[0309] The list display area 1352 is an area for displaying a list of evidence associated with a case.

[0310] The list display area 1352 corresponds to steps S2229 and S2213 in FIG. 22.

[0311] The description confirmation operation unit 1354 is an operation member that accepts an operation for confirming a list of evidence described in document data such as preparatory documents associated with a case.

[0312] The description confirmation operation unit 1354 corresponds to steps S2229 and S2213 in FIG. 22. The server 20, for example, while detecting a description specifying evidence in document data such as preparatory documents, outputs a list of the detected evidence in a list format or the like.

[0313] <7 Operations (Third Embodiment)> The third embodiment will be described. After an expert hears the consultation content from the consulter, organizes the consultation content, clarifies the legal points, and constructs claims corresponding to the points, legal documents such as a complaint and an answer are created. Therefore, in the third embodiment, a technique for the server 20 to support the creation of such documents as a complaint, an answer, and also a contract, a term sheet, etc. will be described. FIG. 25 is a diagram showing the flow of a process for generating legal document data based on the information of a case record.

[0314] In step S2511, the terminal 10 accepts an operation from the user to generate a legal work product. Examples of legal work products include the following. · Those that define the content of an agreement between parties, such as a contract and a term sheet · Those that make legal claims between the parties, such as warning letters and response letters · Those related to legal procedures such as litigation or trial, such as pleadings, written defenses, and preparatory written materials In step S2521, the case management module 2048 of the server 20 extracts information to be included in the work product from the consultation content information associated with the case record in the case management database 216.

[0315] In step S2523, the case management module 2048 of the server 20 uses the extracted information to generate a prompt for causing the work product to be output by the large language model, and transmits the generated prompt to the server 95 of the large language model service.

[0316] In step S2525, the case management module 2048 of the server 20 obtains the result of generating the work product from the server 95 of the large language model service and registers it in the case record of the client in the case management database 216.

[0317] Here, in the case management database 216, at least one of the information of the parties related to the consultation content or the information of the agent in the legal procedure is managed as the case record. In step S2521, the case management module 2048 may extract at least one of these managed information of the parties or the information of the agent. In step S2523, the case management module 2048 may use at least one of the extracted information of the parties or the information of the agent to generate a prompt for outputting at least one of the first document data to be submitted in the litigation procedure, the second document data for notifying the legal claims of the parties between the parties, and the third document data regarding the contract conditions agreed between the parties as the work product. In step S2525, the case management module 2048 receives at least one of the first document data, the second document data, and the third document data as the response output by the server 95 of the large language model service and registers it in the case record of the case management database 216.

[0318] In addition, when the case management module 2048 outputs the first document data in step S2523, it may generate a prompt to output at least one of a complaint, an answer, and a preparatory document, and when outputting the third document data, generate a prompt to output at least one of a contract and a term sheet defining contract terms.

[0319] Here, for a complaint, an answer, and a preparatory document, the standard writing method (format as a document, headings, etc.) is provided in books (which may be books provided by the server 93 of the book browsing service), websites, etc. The case management module 2048 generates a prompt including an instruction to generate a work product by applying the information related to the consultation content while referring to these standard writing methods and the data of sample documents, and provides it to the server 95 of the large language model service, so that the server 95 of the large language model service generates the work product.

[0320] For example, the case management module 2048 may provide the following prompt to the server 95 of the large language model service, so that the server 95 of the large language model service outputs a written statement such as a rebuttal. · There are books (for example, made available for users to view on the server 93 of the book browsing service, etc.) and articles on the web that record examples of written rebuttals for making legal claims. Include in the prompt an instruction to create a written statement that reflects the information on the claimed content stored in the case management database 216 by referring to these examples of writing, for example, examples of rebuttals along the lines of the material facts. · Include in the prompt an instruction to create a written statement in accordance with the guidelines shown in the above-mentioned method of writing a general written rebuttal (for example, separately stating "denial" and "defense"). In addition, to suggest evidence to be presented to supplement the claim along with the counterargument, the case management module 2048 may also provide the following prompts to the server 95 of the large language model service. · Include in the prompt an instruction to enumerate the evidence presented in past cases (records of each case stored in the case management database 216, case laws stored in a case law database, etc., explanations of case laws published as articles or books, etc.) for the defenses asserted in the counterargument. · Include in the prompt an instruction to enumerate the evidence presented in past cases when denying in the counterargument. This enables the user to easily understand, for example, the degree of evidence for which the defense was recognized, and makes it even easier to create the written counterargument.

[0321] Also, in the case management database 216, information on the term sheet that defines the contract terms is managed as a record of the case. In step S2523, the case management module 2048 may generate a prompt to output a contract as the third document data using the information on the term sheet of the case record. The case management module 2048 receives the contract data as the output response from the server 95 of the large language model service.

[0322] Here, for the term sheet and the contract, the standard writing style (document format, headings, etc.) is provided in books (which may be books provided by the server 93 of the book browsing service), websites, etc. The case management module 2048 generates a prompt including an instruction to generate the work product by applying the information (such as contract terms) related to the consultation content while referring to these standard writing styles and the data of sample documents, and provides it to the server 95 of the large language model service, so that the server 95 of the large language model service generates the work product.

[0323] In addition, in step S2523, the case management module 2048 may generate a prompt that includes causing the large language model service server 95 to create a contract definition by referring to a thesaurus or dictionary database that stores the definitions of terms. As a result, a contract with unified terms can be output by the large language model service server 95.

[0324] The case management module 2048 stores the response received from the large language model service server 95 in association with the record of the case from which the information related to the generation of the prompt has been extracted in the case management database 216.

[0325] In step S2527, the server 20 outputs the deliverable information of the case record in the case management database 216 to the terminal 10.

[0326] In step S2513, the terminal 10 displays the information of the generated deliverable.

[0327] FIG. 26 is a diagram showing the flow of a process for generating counterarguments assumed for an argument.

[0328] In step S2611, the terminal 10 receives from the user an operation to generate counterarguments assumed for an argument regarding the argument information stored in association with the case in the case management database 216.

[0329] In step S2621, the case management module 2048 of the server 20 generates a prompt for outputting to the large language model a counterargument against the claim content between the parties in the written data held in the case management database 216.

[0330] Specifically, in the case management database 216, as records of cases, it manages information of written data that describes the content of claims made by the parties for each argument regarding the consultation content. The case management module 2048 generates a prompt that causes the large language model to output counterarguments against the content of claims between the parties in the written data.

[0331] In step S2623, the case management module 2048 of the server 20 transmits the generated prompt to the server 95 of the large language model service. In this way, by providing the generated prompt to the large language model, the case management module 2048 causes the server 95 of the large language model service to output counterarguments.

[0332] In step S2625, the case management module 2048 of the server 20 acquires the information of the output counterarguments from the server 95 of the large language model service and registers it in the case record regarding the counselor in the case management database 216.

[0333] In step S2621, the case management module 2048 may also generate a prompt for searching for counterarguments by using, as search targets, the content of claims in the written data of each case (including other cases different from the case for which counterarguments are to be output) accumulated in the case management database 216 or the data of published judgment documents based on the content of claims between the parties in the written data, and causing the large language model to summarize the search results. In step S2623, the case management module 2048 may cause the summarized counterarguments to be output to the server 95 of the large language model service by providing the generated prompt to the large language model.

[0334] For example, like in the case of a traffic accident or a claim for the return of overpaid money, according to the type of claim in the case (classified by the basis clause), in past cases, the content that the parties have claimed about the argument is identified from the judgment documents and each case managed in the case management database 216, and a prompt for summarizing the content of the identified counterarguments is given to the server 95 of the large language model service, etc., so that the content of the counterarguments assumed from the other party about the argument can be generated by the server 95 of the large language model service.

[0335] In step S2627, the case management module 2048 of the server 20 outputs the counterargument information of the case record in the case management database 216 to the terminal 10.

[0336] Thereby, the case management module 2048 presents the counterargument output by the server 95 of the large language model service to the user.

[0337] In step S2613, the terminal 10 displays the generated counterargument information.

[0338] Figure 27 is a diagram showing the process flow for assisting in the interpretation of a contract.

[0339] In step S2711, the terminal 10 receives from the user an operation of designating a contract in the case record.

[0340] In step S2721, the case management module 2048 of the server 20 generates a prompt for illustrating the citation relationship of each clause of the contract data specified by the user. The prompt includes information for specifying the contract data (for example, a link to the contract data or the full text of the contract data).

[0341] In step S2723, the case management module 2048 of the server 20 transmits the generated prompt to the server 95 of the large language model service.

[0342] In step S2725, the case management module 2048 of the server 20 obtains the diagrammatic information of the citation relationship of each clause output from the server 95 of the large language model service and registers it in the record of the case for the client in the case management database 216.

[0343] In step S2727, the case management module 2048 of the server 20 outputs the diagrammatic information of the citation relationship of each clause of the case record to the terminal 10.

[0344] In step S2713, the terminal 10 displays the generated diagrammatic information of the citation relationship of each clause.

[0345] <8 Screen example (Third embodiment)> FIG. 28 is an example of an operation screen for generating written data related to law.

[0346] The generation support unit 1356 is an area for receiving operations to support the generation of work products.

[0347] The registration information specifying unit 1358 is an operation member for receiving the specification of case information to be included in the work product to be generated.

[0348] In the illustrated example, the registration information specifying unit 1358 is specified to include in the work product information such as the information of the parties and the content of the claims regarding the arguments among the case information. This corresponds to steps S2511, S2521, etc. in FIG. 25.

[0349] The work product specifying unit 1360 is an operation member for receiving the specification of the type of work product to be generated by the server 95 of the large language model service.

[0350] The work product specifying unit 1360 receives the following specifications from the user as the types of work products. · First document data to be submitted in litigation procedures (e.g., complaint, answer, pretrial brief) · Second document data for notifying the legal claims of the parties to each other (e.g., warning letter, response letter) · A third document data which is a document regarding the contract conditions agreed upon between the parties (for example, a contract, a term sheet defining the contract conditions) The dictionary specifying unit 1362 is an operation member that receives the specification of a dictionary to be referred to by the server 95 of the large language model service when generating a work product.

[0351] The generation operation unit 1364 is an operation member that receives an operation to generate a work product by the server 95 of the large language model service based on the specified conditions.

[0352] The generation operation unit 1364 corresponds to step S2511 in FIG. 25 and the like.

[0353] The browsing operation unit 1366 is an operation member that receives an operation to browse the work product generated by the server 95 of the large language model service.

[0354] FIG. 29 is an example of an operation screen that displays assumed counterarguments to an argument.

[0355] The counterargument generation unit 1368 is an area that receives an operation to generate assumed counterarguments from the other party regarding an argument.

[0356] The generation operation unit 1370 is an operation member that receives an operation to generate a counterargument by the server 95 of the large language model service for the content of the claim registered in the record of the case.

[0357] The generation operation unit 1370 corresponds to step S2611 in FIG. 26 and the like.

[0358] The generation result display area 1372 is an area that displays the counterargument generated by the server 95 of the large language model service.

[0359] The generation result display area 1372 corresponds to steps S2627, S2613 in FIG. 26 and the like.

[0360] The case viewing operation unit 1374 is an operation member that accepts an operation to view other cases used as information sources when the server 95 of the large language model service generates a counterargument.

[0361] The case viewing operation unit 1374 may, for example, display links to other cases managed in the case management database 216, information on the claim contents (including the opponent's claim contents) included in the document data registered in other cases.

[0362] The precedent viewing operation unit 1376 is an operation member that accepts an operation to view precedents used as information sources when the server 95 of the large language model service generates a counterargument.

[0363] The precedent viewing operation unit 1376 may, for example, display excerpts of precedents accumulated in the server 91 of the precedent search service, links to precedents, links to court precedents, excerpts of books referring to precedents in the server 93 of the book viewing service, links to the books, etc.

[0364] Figure 30 is an example of an operation screen for assisting in the interpretation of a contract.

[0365] The contract business support unit 1378 is an area that accepts an operation to assist in the operation of referring to and creating a contract based on the data stored in the case management database 216.

[0366] The generation operation unit 1380 is an operation member that accepts an operation to cause the server 95 of the large language model service to generate a contract based on the data registered in the designated case.

[0367] In the illustrated example, the generation operation unit 1380 accepts an operation to organize the description of the contract data registered in the case. This corresponds to step S2711 in Figure 27.

[0368] The generation result display area 1382 is an area that displays the contract generated by the server 95 of the large language model service.

[0369] The server display area 1384 of the information media service is an area for displaying the result output by the server 95 of the large language model service regarding the citation relationship of the contract document.

[0370] In the illustrated example, when there is a citation relationship between clauses, the server display area 1384 of the information media service displays the source clause and the cited clause side by side as a list. This corresponds to steps S2727, S2713, etc. in FIG. 27.

[0371] The dictionary specification unit 1386 is an operation member that accepts a dictionary specification defining the language used in the contract document.

[0372] The template registration operation unit 1388 is an operation member that accepts an operation to register the generated contract document as a template.

[0373] As shown in the illustrated example, the template registration operation unit 1388 displays that it accepts registration as a template in clause units. In this way, using the newly registered or updated template, the server 20 causes the server 95 of the large language model service to generate a contract document according to the user's operation. For example, the server 20 may generate a prompt for creating a contract document according to specified conditions while referring to the templates of the accumulated contract documents, and obtain the generated contract document by sending the prompt to the server 95 of the large language model service. Also, the server 20 may make the registered contract document templates searchable and provide the templates to the user.

[0374] <Modification Example> It may be possible to combine each of the above embodiments. Also, in addition to the aspects described in the above embodiments, the following may be done.

[0375] (1) Improvement in the reliability of the generated answer In the description of the above embodiment, when the answer generated by the server 95 of the large language model service includes the case number, legal provision, source of the guideline, etc. as the basis, whether these generated answers are actual existing cases, laws, guidelines, etc., or the databases of the server 91 of the case search service, the server 92 of the legal search service, the server 93 of the book browsing service, the server 94 of the information media service, etc. may be searched for verification. Specifically, the server 20 determines whether the introduction of the content shown in the case and the case number are included in the answer generated by the server 95 of the large language model service. For example, the case number is assigned according to a predetermined rule (such as "Case No. XXXXX of XX year of Heisei"), and the description conforming to these rules is extracted as the case number. The server 20 inquires of the server 91 of the case search service, etc. based on the case number, obtains the judgment text, abstract, etc. of the judicial precedent, and determines whether it is similar or identical to the result generated by the server 95 of the large language model service. For example, the server 20 may determine similarity based on the degree of word matching included in the text, or may vectorize the meaning of the document and determine similarity based on the vector distance.

[0376] If the server 20 determines that they are not similar, it may cause the server 95 of the large language model service to generate an answer again, or may respond to the user with the case number of the actually existing case as a result of inquiring of the server 91 of the case search service, etc. For example, regarding the introduction text of the case included in the answer generated by the server 95 of the large language model service, a similar judgment text may be searched from the server 91 of the case search service, etc., and the case number of the judicial precedent corresponding to the search result may be responded.

[0377] Thereby, when the answer generated by the server 95 of the large language model service includes a case, etc. that does not actually exist, it is possible to determine whether the generated answer is appropriate by referring to the original case, etc.

[0378] In addition, when the server 20 checks whether the content of the answer generated by the server 95 of the large language model service contains the case number, legal provisions, guideline source, reference documents, etc. as the basis, the server 20 may execute by giving a prompt to the server 95 of the large language model service to perform such verification. After receiving the verification result, the server 20 can evaluate the reliability of the answer generated by the server 95 of the large language model service and may present the evaluation result to the user.

[0379] For example, while presenting the generated answer and the basis to the user, the server 20 may also present to the user whether there is a description in the basis as a result of the verification of the basis.

[0380] Also, for example, if the server 95 of the large language model service outputs that there is no description in the basis as a result of having the server 20 perform such verification, the server 20 may not present the answer generated by the server 95 of the large language model service to the user and may notify the user such as "Failed to generate an answer".

[0381] (2) Specify domestic and overseas information as the information source The server 20 may receive from the user a designation of whether the information is issued domestically or overseas (or in a specific country), and receive an instruction to generate an answer by the server 95 of the large language model service with reference to the information source corresponding to the user's designation (for example, when the user designates Europe, define in the prompt to summarize the results of searching for European information).

[0382] Thereby, in addition to legal collation, information about overseas cases can also be provided to the user.

[0383] (3) Classify questions to generate answers for the large language model service and create prompts according to the classified results In addition to what has been described in the above embodiments, in order to cause the large language model service to generate the answer intended by the user, the user's question may be classified, a prompt may be created according to the classified result, and the created prompt may be given to the server 95 of the large language model service.

[0384] For example, in advance, the server 20 holds a classification list for classifying questions, and for each classification, criteria for causing the large language model service to generate an answer are prepared.

[0385] When the server 20 receives an input of a question from the user, it determines to which classification in the question list the input question belongs. For example, the server 20 uses the user's question as an input and gives a prompt to the effect of "answering which classification in the classification list the input question corresponds to" to the server 95 of the large language model service. Thereby, the server 95 of the large language model service generates an answer as to which classification in the classification list the user's question corresponds to. When the server 20 receives the answer generated by the server 95 of the large language model service, it refers to the classification list, generates a prompt including the "criteria for generating an answer" associated with the corresponding classification, and gives the generated prompt to the server 95 of the large language model service. Thereby, the server 20 can obtain the answer generated by the server 95 of the large language model service for the question received from the user based on the "criteria for generating an answer" shown in the classification list.

[0386] For example, by classifying the user's question, the server 20 may select information resources such as books and case laws as the "criteria for generating an answer".

[0387] For example, when it can be determined from the user's question that the user wants to know about a procedure, it may not be necessary to include case information in the answer to the user. Therefore, in this case, the server 20 creates a prompt with the criterion of "not including cases" for "generating an answer". For example, when the user enters a question such as "Please tell me about the precautions for divorce by mutual consent", it may be desired to provide an answer about the procedure.

[0388] Also, when it can be determined from the user's question that the user wants to know about a case or the definition of a term, it may be possible to generate an answer targeting cases as the "criterion for generating an answer".

[0389] In this way, instead of transmitting all information at once to the server 95 of the large language model service to generate an answer, by giving a prompt for the classification corresponding to the question, it may be possible to suppress the consumption of computing resources of the server 95 of the large language model service.

[0390] (4) Reinvestigation by an expert's instruction when the server 95 of the large language model service conducts an argument and research based on the information in the case record In the above embodiment, with reference to FIGS. 21 to 25, it has been described that by registering the data of the consultation content in the case record, the server 95 of the large language model service extracts legal arguments, conducts a search corresponding to the arguments, and registers the results in the case record.

[0391] The server 20 may present the arguments extracted by the server 95 of the large language model service and the summary of the search results corresponding to the arguments to the user who is an expert, accept the operation of confirming these from the expert, and register them in the case management database 216. At this time, the server 20 may also accept an instruction from an expert such as a lawyer to re-do the extraction of the arguments and the investigation by the search corresponding to the arguments, re-perform the above-mentioned extraction of the arguments, etc., and present the results to the user.

[0392] Regarding the results of legal interpretation, compare with the search results in information sources different from legal-related ones such as legal books and case laws. The server 20 may also present to the user the legal interpretations shown in legal-related books and case laws in a comparable manner to the values of people over time.

[0393] For example, the server 20 searches the servers 94 of information media services and 96 of SNS for legal arguments. For example, when searching SNS etc. for recent past case laws, the reactions of SNS users can be obtained as search results. The server 20 may also present to the user by comparing the interpretation of past case laws with the result summarized by the server 95 of the large language model service from the search of the reactions of users in SNS etc. This can suggest the possibility of changes in future case laws, legal amendments, and changes in academic theories.

[0394] (6) Assistance in reading judgment documents In the above embodiment, an example of searching for legal arguments and corresponding case laws etc. was explained.

[0395] Here, when there is a judgment document of a higher court that modifies and indicates the original judgment, the server 20 may also display the description of the original judgment by reading it again based on the content modified by the judgment document of the higher court. Thereby, for example, by partially modifying the original judgment, the high court judgment can be made easier to read.

[0396] (7) Creating a report on the investigation results regarding legal arguments The server 20 may also cause the server 95 of the large language model service to generate a report in accordance with a predetermined format such as documents and email text to facilitate the expert's reporting of the investigation results based on legal arguments regarding the case to the interviewee.

[0397] (8) Generate a contract while maintaining the indent setting of the contract The server 20 may also set indents in the contract data so that articles, clauses, and items can be identified separately (such as making the amount of indent different for articles, clauses, and items), and generate a contract based on the information of the project record.

[0398] As described above, some embodiments of the present disclosure have been described. However, these embodiments can be implemented in various other forms, and various omissions, replacements, and changes can be made without departing from the gist of the invention. These embodiments and their modifications are intended to be included in the scope and gist of the invention, as well as in the invention described in the claims and its equivalent scope.

[0399] The functions realized by the components described in this specification may be implemented in circuitry or processing circuitry including a general-purpose processor, a specific-purpose processor, an integrated circuit, ASICs (Application Specific Integrated Circuits), a CPU (a Central Processing Unit), a conventional circuit, and / or a combination thereof, which are programmed to realize the described functions. The processor includes transistors and other circuits and is regarded as circuitry or processing circuitry. The processor may be a programmed processor that executes a program stored in a memory.

[0400] In this specification, circuitry, unit, and means are hardware programmed to realize the described functions or hardware that executes them. The hardware may be any hardware disclosed in this specification or any hardware known to be programmed or execute to realize the described functions.

[0401] When the hardware is a processor considered to be of the circuitry type, the circuitry, means, or unit is a combination of hardware and software used to configure the hardware and / or the processor.

[0402] <Supplementary Note> The matters described in the above embodiments are appended below.

[0403] <Supplementary Note of the First Embodiment>

[0404] (Supplementary Note 1) A program for operating a computer equipped with a computer processor, the program comprising steps of: receiving, by the computer processor, input of consultation data indicating the content of a consultation from a consulter; obtaining a summarized result of the consultation content by giving a prompt to a large language model to summarize the consultation content shown in the consultation data along predetermined items; and associating the consultation content summarized along the predetermined items with the consulter and storing it in a storage unit so that an expert responding to the consultation from the consulter can refer to it.

[0405] (Supplementary Note 2) In the receiving step, voice data indicating the consultation content is received as the consultation data, speech-to-text data obtained by speech recognition of the voice data is acquired, and in the step of storing it in the storage unit, the speech-to-text data is stored in association with the consulter, and at least any one of information on the result of determining the speaker, information on the result of determining the emotion, and information on the timing when there was a consultation from the consulter in the speech-to-text data is further stored in the storage unit. The program according to Supplementary Note 1.

[0406] (Supplementary Note 3) In the obtaining step, a prompt for sorting and organizing the consultation content included in the consultation data in chronological order is given as the prompt to obtain a summarized result, and in the step of storing it in the storage unit, the summarized result sorted in chronological order is stored in association with the consulter. The program according to any one of Supplementary Notes 1 to 2.

[0407] (Appendix 4) Based on the results of sorting and summarizing the consultation content by time series using a large language model, each of the sorted results is displayed on the screen as an object, and the user is accepted to perform an operation of rearranging the objects. By giving a prompt to the large language model to generate a text in time series based on the order of rearrangement according to the operation of rearranging the objects, the summarized results sorted by time series are further obtained. The program according to Appendix 3.

[0408] (Appendix 5) In the storage unit, each consultation from the consulter is managed as a case including a case and a predetermined management item. In the step of storing in the storage unit, each item of the consultation content summarized along with the predetermined items is associated with each management item managed as a case in the storage unit and stored. The program according to any one of Appendices 1 to 4.

[0409] (Appendix 6) In the storage unit, at least one of the consultation content from the consulter, the next action to be taken, and the concerns of the consulter is included as a predetermined management item and managed as a case. In the step of obtaining, a prompt is given to the large language model to summarize at least one of the items included in the predetermined management item as a predetermined item, and the summarized result is obtained. The program according to Appendix 5.

[0410] (Appendix 7) In the receiving step, as consultation data, voice data indicating the consultation content is received, the speech-to-text data obtained by speech recognition of the voice data is obtained, and in the step of storing in the storage unit, the speech-to-text data is stored in association with the consulter, and the information on the result of determining the emotion of the consulter is further stored in the storage unit. In the obtaining step, based on the voice data of the consultation content of the consulter and the information on the result of determining the emotion, the concerns of the consulter are extracted, and in the step of storing in the storage unit, the information on the concerns of the consulter is stored as case information. The program according to Appendix 6.

[0411] (Appendix 8) The program further causes a computer processor to execute steps of: receiving an input of image data showing a correlation diagram of persons related to the consultation content from the person seeking consultation; providing a prompt to a large language model to analyze the image data and generate a text showing the correlation of persons, thereby obtaining a text showing the correlation of persons; and outputting the obtained text showing the correlation of persons. The program according to any one of Appendices 1 to 7.

[0412] (Appendix 9) In the receiving step, as consultation data, voice data showing the consultation content is received, and in the step of obtaining the speech-to-text data obtained by speech-to-text conversion of the voice data and storing it in the storage unit, the speech-to-text data is stored in association with the person seeking consultation, and in the speech-to-text data, a word indicating a person is identified. When a word indicating a person is identified, a question asking about the relationship between the identified person and other persons appearing in the consultation content is presented to the person seeking consultation or an expert, and an input of an answer is received. The program further executes the step of storing the received answer in the storage unit.

[0413] (Appendix 10) A method executed by a computer including a computer processor, the method including steps of: the computer processor receiving an input of consultation data showing the consultation content from the person seeking consultation; providing a prompt to a large language model to summarize the consultation content shown in the consultation data along a predetermined item, thereby obtaining a result of summarizing the consultation content; and storing in a storage unit so that an expert responding to the consultation from the person seeking consultation can refer to it, associating the consultation content summarized along the predetermined item with the person seeking consultation.

[0414] (Supplementary Note 11) An information processing apparatus, wherein a control unit of the information processing apparatus executes: a step of receiving an input of consultation data indicating consultation content from a consulter; a step of obtaining a result of summarizing the consultation content by giving a prompt to a large language model so as to summarize the consultation content along predetermined items; and a step of associating the consultation content summarized along the predetermined items with the consulter and storing the same in a storage unit so that an expert responding to the consultation from the consulter can refer to it.

[0415] <Supplementary Note of the Second Embodiment>

[0416] (Supplementary Note 1) A program for operating a computer including a processor, wherein in a storage unit, each of the consultation contents from a consulter is configured to be managed as a record of a case, and the program causes the processor to execute: a step of receiving registration of the consultation content of the consulter and storing the same in the storage unit in association with the record of the case; a step of obtaining information on an argument by giving a prompt for extracting an argument corresponding to the consultation content to a large language model based on the information on the consultation content associated with the record of the case; and a step of storing the obtained information on the argument in the storage unit in association with the record of the case related to the consultation content.

[0417] (Supplementary Note 2) The program according to Supplementary Note 1, wherein in the obtaining step, based on the information on the consultation content, a concern point that the consulter regarding the consultation content is concerned about is specified, and an argument corresponding to the specified concern point is obtained by a large language model.

[0418] (Supplementary Note 3) The program causes the processor to further perform a step of searching a plurality of information sources related to laws, such as a legal database, a case law database, a legal book database, etc., for the arguments stored in the storage unit in association with the records of the case; a step of summarizing the results of searching the plurality of information sources by a large language model; and a step of storing the summarized results in the storage unit in association with the records of the case related to the arguments. The program is as described in any one of Appendices 1 to 2.

[0419] (Appendix 4) In the searching step, search the case law database, detect in the searched cases that the appellate court has amended the judgment of the lower court, and present the content of the amendment of the judgment amended by the appellate court related to the detection to the user. The program is as described in Appendix 3.

[0420] (Appendix 5) In the searching step, for the information stored in a plurality of information sources, search for a first information source that is past information in chronological order and a second information source that is future information in chronological order compared to the first information source, compare the results of searching the first information source and the results of searching the second information source, output the different content to a large language model, and present the output result of the different content to the user. The program is as described in Appendix 3.

[0421] (Appendix 6) In the storage unit, manage materials, evidence, preparatory documents, and other written data used in litigation procedures related to the consultation content of the case in association with the records of the case, and generate information indicating a list of these based on various written data registered in the storage unit by a large language model. The program is as described in any one of Appendices 1 to 5.

[0422] (Appendix 7) In a storage unit, in association with a case record, materials, evidence, preparatory documents, and other written data used in litigation procedures related to the consultation content of the case are managed. Based on the data of the preparatory documents registered in the storage unit, the evidence described in the preparatory documents is extracted, and the extracted result is presented to the user. The program described in Supplementary Note 6.

[0423] (Supplementary Note 8) A method executed by a computer including a processor and a storage unit, wherein the storage unit is configured to manage each consultation content from a counselor as a case record. The method includes steps in which the processor receives registration of the consultation content from the counselor and stores it in the storage unit in association with the case record, and gives a prompt for extracting arguments corresponding to the consultation content to a large language model based on the information of the consultation content associated with the case record, thereby obtaining argument information, and storing the obtained argument information in the storage unit in association with the case record related to the consultation content.

[0424] (Supplementary Note 9) An information processing apparatus including a storage unit, wherein the storage unit is configured to manage each consultation content from a counselor as a case record, and the control unit of the information processing apparatus receives registration of the consultation content from the counselor and stores it in the storage unit in association with the case record, and gives a prompt for extracting arguments corresponding to the consultation content to a large language model based on the information of the consultation content associated with the case record, thereby obtaining argument information, and storing the obtained argument information in the storage unit in association with the case record related to the consultation content.

[0425] (Supplementary Note of the Third Embodiment)

[0426] (Supplementary Note 1) A program for operating a computer equipped with a processor, configured to manage each piece of consultation content from a consulter as a record of a case in a storage unit, the program causes the processor to: extract information to be included in a work product from information of the consultation content associated with the record of the case; generate a prompt for causing the work product to be output by a large language model using the extracted information; provide the generated prompt to the large language model and receive a response from the large language model; and present the work product as the received response to the user.

[0427] (Appendix 2) In the storage unit, at least one of information on the parties involved in the consultation content or information on an agent in a legal procedure is managed as a record of the case. In the extraction step, at least one of the information on the parties or the information on the agent is extracted. In the generation step, at least one of the extracted information on the parties or the information on the agent is used to generate a prompt for outputting at least one of first document data to be submitted in a litigation procedure, second document data for notifying the parties of the legal claims of the parties, and third document data which is a document regarding the contract terms agreed upon between the parties as a work product. In the receiving step, at least one of the first document data, the second document data, and the third document data is received as the output response. The program according to Appendix 1.

[0428] (Appendix 3) In the generation step, when outputting the first document data, a prompt for outputting at least one of a complaint, an answer, and a pretrial brief is generated. When outputting the third document data, a prompt for outputting at least one of a contract and a term sheet defining the contract terms is generated. The program according to Appendix 2.

[0429] (Appendix 4) In the storage unit, as a record of a case, it manages information of a term sheet that defines contract terms. In the generating step, it generates a prompt for outputting a contract as third document data using the information of the term sheet. In the receiving step, it receives the contract data as the output response. A program according to any one of Appendices 2 to 3.

[0430] (Appendix 5) A program according to Appendix 4, which in the generating step generates a prompt including creating a definition of a contract by referring to a dictionary or a dictionary database that records definitions of terms.

[0431] (Appendix 6) In the storage unit, as a record of a case, it manages information of written data that records the content of claims made by parties for each point of argument regarding the consultation content. The program further generates, for the processor, a prompt for outputting a counterargument against the content of claims between parties in the written data to a large language model, gives the generated prompt to the large language model to output a counterargument from the large language model, and presents the output counterargument to the user. A program according to any one of Appendices 1 to 5.

[0432] (Appendix 7) Based on the content of claims between parties in the written data, it searches for the content of claims in the written data of each case stored in the storage unit or the data of published judgment documents as search targets for counterarguments, generates a prompt for summarizing the search results by the large language model, and gives the generated prompt to the large language model to output a summarized counterargument from the large language model. A program according to Appendix 6.

[0433] (Appendix 8) In the storage unit, contract data is managed in association with the records of cases. In the generating step, a prompt for illustrating the citation relationships of each clause of the contract data is generated. In the receiving step, the illustrated information is received by providing the prompt to a large language model. A program according to any one of Appendices 1 to 7.

[0434] (Appendix 9) A program according to any one of Appendices 1 to 8, wherein the response from the large language model received in the receiving step is stored in the storage unit in association with the record of the case from which the information related to the generation of the prompt has been extracted.

[0435] (Appendix 10) A method for operating a computer equipped with a processor, wherein in the storage unit, each consultation content from a consultant is configured to be managed as a record of a case, and the method includes steps in which the processor extracts information to be included in the work product from the information of the consultation content associated with the record of the case, generates a prompt for outputting the work product by a large language model using the extracted information, provides the generated prompt to the large language model, receives a response from the large language model, and presents the work product as the received response to the user.

[0436] (Appendix 11) An information processing apparatus, wherein in the storage unit, each consultation content from a consultant is configured to be managed as a record of a case, and the control unit of the information processing apparatus extracts information to be included in the work product from the information of the consultation content associated with the record of the case, generates a prompt for outputting the work product by a large language model using the extracted information, provides the generated prompt to the large language model, receives a response from the large language model, and presents the work product as the received response to the user.

Claims

1. A program for operating a computer having a computer processor, the program causing the computer processor to: A step of receiving input of consultation data indicating the contents of the consultation from a person who seeks advice; Prompting a large-scale language model to summarize the consultation contents indicated in the consultation data along predetermined items, thereby obtaining a result summarizing the consultation contents; and associating the consultation content summarized along the predetermined items with the client, and storing the summary in a storage unit so that an expert who responds to the consultation from the client can refer to the summary.

2. In the receiving step, voice data indicating the consultation content is received as the consultation data, obtaining transcription data obtained by transcribing the audio data; In the step of storing in the storage unit, the transcription data is stored in association with the client; The program according to claim 1 , further storing in a memory unit at least one of information on the results of determining the speaker in the transcription data, information on the results of determining emotions, and information on the timing of the consultation from the client.

3. In the acquiring step, the prompt is given to organize the consultation contents included in the consultation data in a chronological order, thereby acquiring the summarized result; 2. The program according to claim 1, wherein in said step of storing in said storage unit, said summarized results organized in chronological order are stored in association with said client.

4. based on the result of dividing and organizing the consultation contents in a chronological order by the large-scale language model, the divided results are displayed on a screen as objects, and an operation of rearranging the objects is accepted from a user; 4. The program of claim 3, further comprising: providing the large-scale language model with prompts to generate sentences in chronological order based on the rearranged order in response to an operation of rearranging the objects, thereby obtaining the summarized results organized in chronological order.

5. In the storage unit, each consultation from a client is managed as a case including predetermined management items, 2. The program according to claim 1, wherein in the step of storing in the memory unit, each item of the consultation content summarized along the specified items is stored in association with each management item managed as a case in the memory unit.

6. In the storage unit, as the predetermined management items, The contents of the consultation from the consultant, Next action is the next step to be taken. The consultant's concerns, The case includes at least one of the above, and is managed as the case.

6. The program according to claim 5, wherein in the obtaining step, the summarized result is obtained by providing the prompt to the large-scale language model to summarize at least one of the specified management items as the specified item.

7. In the receiving step, voice data indicating the consultation content is received as the consultation data, obtaining transcription data obtained by transcribing the audio data; In the step of storing in the storage unit, the transcription data is stored in association with the client; The information on the result of judging the emotion of the client is further stored in a storage unit. In the step of acquiring, a concern of the client is extracted based on voice data of the consultation content of the client and information of the result of the emotion determination, 7. The program according to claim 6, wherein in said storing in said storage unit, information on the concerns of said client is stored as information on said case.

8. The program further causes the computer processor to: A step of receiving an input of image data showing a correlation diagram of people related to the consultation content from a person who is seeking advice; obtaining person correlation sentences by analyzing the image data and prompting a large scale language model to generate person correlation sentences; The program according to claim 1 , further comprising: a step of outputting a sentence indicating the obtained correlation of the persons.

9. In the receiving step, voice data indicating the consultation content is received as the consultation data, obtaining transcription data obtained by transcribing the audio data; In the step of storing in the storage unit, the transcription data is stored in association with the client; identifying words indicative of people in the transcription data; a step of presenting a question to a client or an expert asking about a relationship between the identified person and other characters in the consultation content when the word indicating the person is identified, and receiving an input of a response; The program according to claim 1 , further comprising the step of: storing the received response in the storage unit.

10. 11. A computer-implemented method comprising a computer processor, the method comprising: A step of receiving input of consultation data indicating the contents of the consultation from a person who seeks advice; Prompting a large-scale language model to summarize the consultation contents indicated in the consultation data along predetermined items, thereby obtaining a result summarizing the consultation contents; and associating the consultation content summarized along the predetermined items with the client and storing the summary in a storage unit so that it can be referenced by a specialist who responds to the consultation from the client.

11. An information processing device, comprising: A step of receiving input of consultation data indicating the contents of the consultation from a person who seeks advice; Prompting a large-scale language model to summarize the consultation contents indicated in the consultation data along predetermined items, thereby obtaining a result summarizing the consultation contents; and storing the consultation content summarized along the predetermined items in a storage unit so as to be referenceable by a specialist who responds to the consultation from the client, in association with the consultation content summarized along the predetermined items and the client.

Images