Information processing method and information processing device

Abstract

An information processing method according to an embodiment of the present invention comprises: inputting character data related to a legal case into a trained model; and outputting a case overview which is generated by the trained model on the basis of the character data and which relates to at least the past history of the legal case and matters pertaining to an examination of a legal document related to the legal case.

Description

Information processing method and information processing device

[0001] An embodiment of the present disclosure relates to an information processing method and an information processing device.

[0002] In the past, a business information management system has been proposed that aims to effectively utilize information related to legal business, and that includes a contract consultation database that records contract negotiation request information requesting negotiations to conclude a contract and contract request detail information indicating the details of the request to conclude a contract, a contract management database that records concluded contract information indicating matters related to the concluded contract, and an information management server that, when an instruction to the effect that a contract has been concluded is input from a terminal device, creates concluded contract information based on the contract negotiation request information and contract request detail information recorded in the contract consultation database, records this in the contract management database, and outputs it to the terminal device in response to an instruction from the terminal device. However, there is an improvement in business efficiency regarding the use of information related to legal business.

[0003] JP 2001-216407 A

[0004] The problem to be solved by the present disclosure is to provide a support technology that can increase the efficiency of using information related to legal work.

[0005] An information processing method according to an embodiment includes inputting character data related to a legal case into a trained model, and outputting a case summary generated by the trained model based on the character data, the case summary relating to at least the history of the legal case and matters related to the review of legal documents related to the legal case.

[0006] FIG. 1 is a schematic diagram illustrating an overview of an information processing system according to a first embodiment. FIG. 2 is a block diagram illustrating an example of the hardware configuration of an information processing device according to the first embodiment. FIG. 3 is a block diagram illustrating an example of the hardware configuration of a terminal device according to the first embodiment. FIG. 4 is a diagram illustrating an example of functional blocks of an information processing device according to the first embodiment. FIG. 5 is an example of a prompt for generating a "summary" according to the first embodiment. FIG. 6 is an example of a "summary" generated by the prompt of FIG. 5. FIG. 7 is another example of a "summary" generated by the prompt of FIG. 5. FIG. 8 is an example of a prompt for generating a "topic" according to the first embodiment. FIG. 9 is an example of a "topic" generated by the prompt of FIG. 8. FIG. 10 is another example of a "topic" generated by the prompt of FIG. 8. FIG. 11 is a diagram illustrating an example of a case display screen according to the first embodiment. FIG. 12 is a diagram illustrating an example of a case summary editing screen according to the first embodiment. FIG. 13 is a diagram illustrating an example of a display mode of a case summary of a legal case according to the first embodiment. FIG. 14 is a diagram illustrating an example of functional blocks of a terminal device according to the first embodiment. Fig. 15 is a flowchart showing an example of the flow of a process for generating a case summary for a legal case according to the first embodiment. Fig. 16 is a flowchart showing an example of the flow of a process for automatically generating a case summary for a legal case according to the second embodiment. Fig. 17 is a flowchart showing an example of the flow of a process for correcting a case summary for a legal case according to the second embodiment. Fig. 18 is a flowchart showing an example of the flow of a process for automatically updating a case summary for a legal case according to the second embodiment.

[0007] (First embodiment) Hereinafter, an embodiment will be described in detail with reference to the drawings. In the following description, components having substantially the same functions and configurations are denoted by the same reference numerals, and explanations will be provided only when necessary. In addition, the embodiment can be combined with other embodiments, modifications, and conventional techniques.

[0008] Fig. 1 is a schematic diagram for explaining an overview of an information processing system S according to this embodiment. As shown in Fig. 1, the information processing system S according to this embodiment includes, for example, an information processing device 10 according to this embodiment, and at least one terminal device T1 to TN that can communicate with the information processing device 10 via a network N. The terminal devices T1 to TN are operated by, for example, users P1 to PN, respectively. Note that the configuration of the terminal devices T1 to TN will be described as an example.

[0009] Here, the information processing system S according to this embodiment is, for example, a system that supports the creation of a case summary of a legal case on a browser of a terminal device T1.

[0010] As used herein, a legal case refers to a unit of legal work, such as a contract or legal document requested of a company's legal staff. A legal case may be, for example, a request or a contract. It may also be a group of conversations or arbitrary data, such as a social networking service (SNS) channel or thread, an email thread, or a case registered in a system. A legal case may also be determined by rules, such as the regulations and policies of the company to which the user belongs, the individual user, or the settings of the system vendor. Preferably, each legal case can be identified by, for example, a case number, case name, or client name. Hereinafter, information that can identify a legal case will be simply referred to as case information.

[0011] In this embodiment, a case summary of a legal case includes at least two sentences: a "summary" and a "topic." The "summary" is the history of the legal case, particularly the past history. The "topic" is a matter related to the review, including changes to legal documents related to the legal case. Specific examples of a "summary" and a "topic" will be described later. The "summary" is an example of a first sentence in this embodiment. The "topic" is an example of a second sentence in this embodiment. In this specification, a "sentence" may include not only a sentence consisting of multiple sentences, but also very short information such as only one word, only one sentence, or a list of bullet points.

[0012] In this specification, a document refers to a document in any language that is created in accordance with certain rules or standards, such as a regulation, law, or custom. Furthermore, in this specification, a legal document includes contracts, terms and conditions, and other legal documents in general. Furthermore, legal documents handled by the information processing system S may be written in any language.

[0013] The information processing system S of this embodiment receives input of character data related to a legal case, and generates and outputs a case summary of the legal case from the character data.

[0014] The text data may be any data containing text information, and may be in any format. In this embodiment, the text data input to the information processing system S is, for example, the body of an email. The text data may also be, for example, posts to an information processing system, posts on an internal bulletin board, posts using a chat tool such as Slack (registered trademark), documents created using editing software, or other types of text data. To generate an appropriate case summary, the text data is preferably data containing negotiations with the counterparty of a contract. Because email is often used for negotiations with counterparties, emails or text data incorporating email information are likely to contain descriptions of past history and the purpose of changes, making them preferable as source text for summarization. Furthermore, emails, in particular, often contain more text not directly related to the content of the case, such as greetings and signatures, than other tools. Therefore, when a person in charge later reviews the text to understand the history of a legal case, the amount of text is often large compared to the amount of information obtained, often resulting in a heavy review workload. For this reason, summarizing emails as case summaries can significantly contribute to improving work efficiency. Note that, in this specification, text data also includes tabular data. Additionally, the term "character data" in this specification also includes text in comments in text files generated by word processing software. Specifically, text in legal documents (e.g., contracts) exchanged with other parties or files related to other legal cases, as well as comments in files created using functions such as Microsoft Word (registered trademark), can also be used as character data to generate a case summary. Note that in this embodiment, "email" refers not only to data formats specifically for e-mail (e.g., data with the extension ".eml"), but also to text data containing the contents of e-mail and text data containing portions of the contents of e-mail.

[0015] In this embodiment, the terminal device T1 is a client device managed by a user P1 who uses a legal case summary creation support service. The user P1 is, for example, a legal affairs officer. The terminal device T1 displays a legal case summary including a "summary" and a "topic" on the screen of a display device, for example, based on information from the information processing device 10.

[0016] The information processing system S may have functions other than supporting the creation of legal case summaries. For example, the information processing device 10 of the information processing system S may have a character data storage function, such as storing and managing legal documents or managing the sending and receiving of emails. In this case, users P1 to PN may access the information processing device 10 from terminal devices T1 to TN via the network N, store legal documents in the information processing device 10, send and receive emails using functions provided by the information processing device 10, exchange messages using a chat function within the system, or exchange messages using a chat tool linked to the system, thereby storing character data within the system.

[0017] The information processing system S according to this embodiment may also include a document creation support function and a document viewing function. The document creation support function is a function for editing (correcting, adding, deleting, etc.) the content of legal documents and for providing information to support the editing. The information processing system S according to this embodiment may also use information related to the editing of legal documents (e.g., a revision history of wording changes made to contracts and comments added to documents) as text data, for example, to generate "topics." The document viewing function is a function for supporting the viewing (display) of documents on a terminal device.

[0018] The services (document viewing support services) provided by the document viewing function of the information processing system S of this embodiment may include, for example, at least one of the following: performing a document review process on a legal document on a web browser and providing the results to the user; performing document editing in response to instructions from the user; saving the edited document or providing it to the user; editing additional information associated with the legal document (e.g., meta information, comment information, chat information, format information, etc.); creating a new document (including copies); comparing multiple documents (e.g., clearly indicating difference information); automatically proofreading legal documents; and performing various processes using the translation display function.

[0019] In this specification, a document review process refers to, for example, a process of analyzing or evaluating information contained in a legal document based on certain criteria and outputting the results. For example, a document review process includes checking and examining whether the information contained in a legal document is appropriate based on certain criteria. Specifically, the process includes, for example, comparing the wording in the document with a pre-prepared checklist and outputting the comparison results or the issues contained in the checklist. Furthermore, editing refers to, for example, adding, deleting, or changing information contained in the edited object.

[0020] Next, the specific configurations of the information processing device 10 and the terminal device T1 of this embodiment will be described.

[0021] (Hardware Configuration of Information Processing Device 10) FIG. 2 is a block diagram showing an example of the hardware configuration of the information processing device 10 according to an embodiment. The information processing device 10 according to this embodiment is connected via communication with terminal devices T1 to TN and performs various processes in response to requests from the terminal devices T1 to TN. The information processing device 10 is managed, for example, by a provider that provides a legal case summary creation support service. For example, the information processing device 10 provides the legal case summary creation support service to the terminal device T1 in response to an instruction to create a legal case summary received via the network N from the terminal device T1 as a client device.

[0022] As shown in FIG. 2, the information processing device 10 of this embodiment is a computer, and includes, as an example, a processor 11, a main storage device (memory) 13, an auxiliary storage device (storage) 15, a network interface 17, a device interface 18, and a bus 19 connecting each component.

[0023] The information processing device 10 of this embodiment shown in Figure 2 includes one of each component, but may also include multiple of the same component. Also, while Figure 2 shows one information processing device 10, the software may be installed on multiple computers, with each of the multiple computers executing the same or different parts of the software. In this case, a form of distributed computing may be used in which each computer communicates with the other computers via a network interface 17 or the like to execute the processing. In other words, the information processing device 10 of this embodiment may be configured as a system that realizes various functions described below by having one or more computers execute instructions stored in one or more storage devices.

[0024] Various calculations of the information processing device 10 of this embodiment may be executed in parallel using one or more processors 11, or using multiple computers via a network. Furthermore, various calculations may be distributed to multiple processing cores within the processor 11 and executed in parallel. Furthermore, some or all of the processes, means, etc. disclosed herein may be executed by at least one of a processor and a storage device provided on a cloud that can communicate with the information processing device 10 via a network. Thus, the processing of the information processing system S of this embodiment may be executed in the form of parallel computing using one or more computers.

[0025] The processor 11 may be an electronic circuit (such as a processing circuit, processing circuitry, CPU (Central Processing Unit), GPU (Graphics Processing Unit), FPGA (Field Programmable Gate Array), or ASIC (Application Specific Integrated Circuit)) including a control device and an arithmetic device of the information processing device 10. The processor 11 may also be a semiconductor device including a dedicated processing circuit. The processor 11 is not limited to an electronic circuit using electronic logic elements, and may be realized by an optical circuit using optical logic elements. The processor 11 may also include an arithmetic function based on quantum computing.

[0026] The processor 11 performs arithmetic processing based on data and software (programs) input from each device, etc., configured internally of the information processing device 10, and can output arithmetic results and control signals to each device, etc. The processor 11 may control each component constituting the information processing device 10 by executing an OS (Operating System), applications, etc. of the information processing device 10.

[0027] The information processing device 10 in this embodiment may be realized by one or more processors 11. Here, the processor 11 may refer to one or more electronic circuits arranged on one chip, or may refer to one or more electronic circuits arranged on two or more chips or two or more devices. When multiple electronic circuits are used, the electronic circuits may communicate with each other via wire or wirelessly.

[0028] The main memory device 13 in this embodiment is a storage device that stores instructions executed by the processor 11 and various data, and information stored in the main memory device 13 is read by the processor 11. The auxiliary memory device 15 is a storage device other than the main memory device 13 and may be composed of a storage medium such as an HDD (Hard Disk Drive) or flash memory. Note that these storage devices refer to any electronic component capable of storing electronic information and may be semiconductor memory. The semiconductor memory may be either volatile memory or nonvolatile memory. The storage device for saving various data used in the information processing device 10 in this embodiment may be realized by the main memory device 13 or the auxiliary memory device 15, or may be realized by an internal memory built into the processor 11. For example, the main memory device 13 and the auxiliary memory device 15 may be collectively referred to as a storage unit or a storage device.

[0029] If the information processing device 10 has functions such as storing and managing legal documents or managing the sending and receiving of e-mails, the auxiliary storage device 15 may store, for example, legal documents sent from terminal devices T1 to TN via network N, or e-mails sent and received in the past.

[0030] The auxiliary storage device 15 may also store a trained model capable of summarizing character data. The trained model of this embodiment is a trained model known as generative AI (Artificial Intelligence), such as a language model or a large-scale language model (LLM: Large Language Models). Large-scale language models include, for example, BERT (Bidirectional Encoder Representations from Transformers), GPT (Generative Pre-trained Transformer), and in particular ChatGPT or GPT-4. The trained model may exist outside the information processing device 10. For example, the trained model may exist outside the information processing device 10, for example, on an external server connected to the information processing device 10 via a network N. In this case, the case summary generation unit 113 inputs character data to a trained model on an external server via the network N, and acquires output from the trained model via the network N. The trained model may also be present on a server device or the like external to the information processing system S.

[0031] Multiple processors 11 may be connected (coupled) to one storage device (main storage device 13 and / or auxiliary storage device 15), or a single processor 11 may be connected. Multiple storage devices may be connected (coupled) to one processor 11. When the information processing device 10 in this embodiment is configured with at least one storage device and multiple processors connected (coupled) to this at least one storage device, it may include a configuration in which at least one of the multiple processors is connected (coupled) to at least one storage device. This configuration may also be realized by storage devices and processors 11 included in multiple computers. Furthermore, it may include a configuration in which a storage device is integrated with the processor 11 (for example, a cache memory including an L1 cache and an L2 cache).

[0032] The network interface 17 of this embodiment is an interface for connecting to the network N wirelessly or via a wire. The network interface 17 may be an appropriate interface, such as one conforming to an existing communication standard. Information may be exchanged with an external device connected via the network N through the network interface 17. The network N may be a wide area network (WAN), a local area network (LAN), a personal area network (PAN), or a combination thereof, as long as information is exchanged between the information processing device 10 and the external device. An example of a WAN is the Internet, an example of a LAN is IEEE 802.11 or Ethernet (registered trademark), and an example of a PAN is Bluetooth (registered trademark) or NFC (Near Field Communication), etc.

[0033] The device interface 18 of this embodiment is an interface such as a USB (Universal Serial Bus) that directly connects to an output device such as a display device, an input device, or an external device. The external device may also be a storage device (memory), network storage, or HDD. The external device may also be a device that has some of the functions of the components of the information processing device 10 of this embodiment. In other words, the information processing device 10 may transmit or receive some or all of the processing results of the external device. The display device connected to the device interface 18 may be, for example, a liquid crystal display (LCD), a cathode ray tube (CRT), a plasma display panel (PDP), or an organic electroluminescence (EL) panel. Various input interfaces, such as a keyboard, a mouse, or a touchpad, can be used as the input device connected to the device interface 18. Furthermore, the information processing device 10 may be configured to include an output device such as a display device and an input device.

[0034] (Hardware Configuration of Terminal Device T1) Fig. 3 is a block diagram showing an example of the hardware configuration of terminal device T1 according to this embodiment. In a legal case summary creation support service, terminal device T1 of this embodiment displays various operation screens generated by information processing device 10 and a legal case summary. In addition, terminal device T1 transmits information related to operations by user P1 to information processing device 10.

[0035] 3, the terminal device T1 is a computer, and includes, as an example, a processor 21, a main storage device (memory) 23, an auxiliary storage device (storage) 25, a network interface 27, a device interface 28, a bus 29 connecting the respective components, a display device 30 connected to the device interface 28, and an input device 31 connected to the device interface 28. Note that the specific hardware configurations of the processor 21, the main storage device 23, the auxiliary storage device 25, the network interface 27, the device interface 28, and the bus 29 are similar to those of the processor 11, the main storage device 13, the auxiliary storage device 15, the network interface 17, the device interface 18, and the bus 19, respectively, and therefore description thereof will be omitted. Specific examples of the terminal device T1 include, but are not limited to, a PC (Personal Computer) or a tablet terminal.

[0036] The display device 30 is a display device such as an LCD, a CRT, a PDP, an organic EL panel, etc. Based on instructions from the processor 21, the display device 30 displays, for example, a case display screen and a case summary editing screen, which will be described later.

[0037] The input device 31 is a device such as a keyboard, a mouse, or a touch panel. The user P1 can input various editing instructions, for example, in a document editing process, using the input device 31. The instructions input from the input device 31 are sent to the processor 21.

[0038] The terminal device T1 may also include other external devices (for example, a camera, a microphone, a motion capture device, various sensors, etc.) connected via the device interface 18 in order to manipulate and edit document data.

[0039] (Functional Configuration of Information Processing Device 10) Next, the functional configuration of the information processing device 10 will be described. FIG. 4 is a diagram showing an example of functional blocks of the information processing device 10 according to this embodiment. The information processing device 10 includes, for example, an acquisition unit 111, an operation acceptance unit 112, a case summary generation unit 113, a display control unit 114, and a registration processing unit 115 as functions realized by the processor 11. The functions realized by the acquisition unit 111, the operation acceptance unit 112, the case summary generation unit 113, the display control unit 114, and the registration processing unit 115 are each stored as a program in, for example, the main storage device 13 or the auxiliary storage device 15. The processor 11 can realize various functions (the acquisition unit 111, the operation acceptance unit 112, the case summary generation unit 113, the display control unit 114, and the registration processing unit 115) by reading and executing the program stored in the main storage device 13 or the auxiliary storage device 15.

[0040] The acquisition unit 111 of this embodiment acquires character data related to a legal case. As described above, the character data in this embodiment may be data containing character information, and may be in any format.

[0041] Regarding the method of acquiring character data, for example, if character data is stored in advance in the auxiliary storage device 15 of the information processing device 10, the acquisition unit 111 may search for and acquire only character data related to the legal case for which a case summary is to be generated from the stored character data using, for example, case information about the legal case for which a case summary is to be generated as a search condition. Alternatively, the acquisition unit 111 may acquire character data from the terminal device T1 via the network N. For example, the display control unit 114 (described later) may display a character data input field on a browser on the display device of the terminal device T1, and the user P1 may input character data into the character data input field. Alternatively, the terminal device T1 may save an electronic file of character data via the network N in a default folder provided in the auxiliary storage device 15 of the information processing device 10, in response to an operation by the user P1. Alternatively, the acquisition unit 111 may receive an email or SMS (Short Message Service) message via the network N and acquire the character data from the message.

[0042] The operation reception unit 112 of this embodiment receives an operation by the user P1 from the terminal device T1 via the network N. For example, when the user P1 presses an image button indicating the generation of a case summary, which is provided on the browser by the display control unit 114 described below, the operation reception unit 112 receives the operation by the user P1.

[0043] The case summary generation unit 113 of this embodiment inputs character data related to a legal case into the trained model and causes the trained model to output a case summary of the legal case. Specifically, the case summary generation unit 113 inputs the character data acquired by the acquisition unit 111 into the large-scale language model. The large-scale language model of this embodiment may be stored in, for example, a server device outside the information processing system S.

[0044] As described above, in this embodiment, the case summary of a legal case includes at least two sentences: a “summary” and a “topic.” Therefore, the case summary generation unit 113 of this embodiment inputs, to the trained model, character data and a prompt instructing the trained model to generate two sentences: a “summary” and a “topic” from the character data.

[0045] FIG. 5 is an example of a prompt for generating a "summary" according to this embodiment. As described above, the "summary" is the historical background of a legal case. The prompt for generating a "summary" is an example of a first prompt according to this embodiment. Note that the prompt shown in FIG. 5 may be disclosed to user P, or may be used only via the system without being disclosed.

[0046] As shown in FIG. 5 , the prompt for generating a “summary” in this embodiment includes multiple categories included in the “summary” and at least one example for each of the multiple categories. Examples of the multiple categories are “purpose,” “request summary,” “approach,” and “related contracts.” In the example shown in FIG. 5 , the prompt for generating a “summary” also includes sentences explaining each of the multiple categories. The explanation for “purpose” is “the purpose of the person who requested the review or creation of a contract or memorandum.” The explanation for “request summary” is “the content requested by the person who requested the review or creation of a contract or memorandum.” The explanation for “approach” is “what was done by the legal department member who was requested to review or create the contract or memorandum.” The explanation for “related contracts” is “contracts, memoranda, or concluded contracts that may be related to the request.” Note that although the category is called “related contracts,” it may also include legal documents other than contracts. By including topics (categories) and examples to be extracted in this way, “summary” data can be appropriately obtained using a large-scale language model.

[0047] An example of "purpose" is "changing the service usage price." An example of "request summary" is "there was a price error in the service usage contract concluded with the other party, so we will create a memorandum to correct this." An example of "approach" is "the service usage price was changed by memorandum." Examples of "related contracts" are "concluded service usage contract" and "reference memorandum data."

[0048] Furthermore, the prompt for generating a "summary" in this embodiment includes a sentence that explains the input character data to the large-scale language model. The explanation written before the input of character data in the prompt is, for example, "You are an expert in compiling XX in the legal department. Please extract information from the XX below." The "XX" indicates the type of character data to be input. In the example shown in FIG. 5 , the text of an email is input as text data after the phrase, "You are an expert in compiling request emails from the legal department. Please extract information from the email below." For example, the case summary generation unit 113 inputs text data, such as the text of an email acquired by the acquisition unit 111, into the input field, thereby inputting the text data into the trained model. Note that one or more emails may be input into the trained model. For example, if emails sent and received in the past are previously associated with case information and stored in the auxiliary storage device 15, the case summary generation unit 113 inputs the text of multiple emails associated with the case information for which a case summary is to be generated into the trained model. Furthermore, the character data may include not only the body of the email but also attached files. The description indicating the type of character data in the prompt may be changeable by the user P. If the prompt is not disclosed to the user P, for example, the display control unit 114 (described later) may display a screen on the browser of the terminal device T1 where the "type of character data" and the "character data" can be input.

[0049] If the number of characters in the character data related to the case for which a case summary is to be generated exceeds the number of characters that can be input to the trained model at one time, the case summary generation unit 113 may divide the character data into portions of the number of characters that can be input to the trained model at one time, input the divided portions to the trained model, and then input a prompt to the trained model instructing it to integrate the input character data. Also, the number of characters or data size of character data such as summaries may be reduced in advance using a different or the same trained model, and then the data may be input to the trained model to obtain a case summary of the legal case.

[0050] The categories, examples, and wording before the input of character data shown in FIG. 5 are all examples. For example, the categories and examples may be customized to differ for each user P1-PN to whom the legal case summary creation support service is provided or for each organization to which the user belongs. This allows a legal case summary tailored to the user's needs. Furthermore, when text other than the body of an email is input as character data, multiple words may be prepared to be written before the input of character data, depending on the format of the text data. For example, the wording "request email" in the above text may be changed to "request."

[0051] Fig. 6 is an example of a "summary" generated by the prompt in Fig. 5. In this embodiment, the case summary generation unit 113 inputs character data and the prompt for generating a "summary" shown in Fig. 5 into the trained model, causing the trained model to output the "summary" shown in Fig. 6. As shown in Fig. 6, the "summary" output from the trained model lists sentences generated based on document data for each of multiple categories included in the prompt for generating a "summary."

[0052] For example, in the example shown in FIG. 6, the category "Purpose" states "Correction of a price error in the contract with A Co., Ltd." The category "Request Summary" states "There was a price error in the contract with A Co., Ltd., so we were requested to conclude a memorandum." The category "Approach" states "A memorandum was concluded to correct the price error." The category "Related Contracts" states the names of documents such as "Concluded Contracts" and "Reference Prices."

[0053] The "related contracts" may include a link destination, such as a URL (Uniform Resource Locator) for storing a contract related to the target legal case, or the file name of the contract. If there are multiple related contracts, the trained model may output the link destinations and contract file names, for example, by contract type or document type. If there are multiple versions of a contract related to the target legal case, the "related contracts" may display only the latest version, or may display two different versions, for example, the first version and the latest version. As described above, the "related contracts" may include documents related to the legal case for which a case summary is to be generated, such as reference information on laws, regulations, guidelines, etc., in addition to contracts. Price lists, specifications, etc. may also be listed as "related information" in the same way as the "related contracts."

[0054] Figure 7 is another example of a "summary" generated by the prompt in Figure 5. Figure 7 shows a "summary" output from a trained model to which character data different from that in the example shown in Figure 6 has been input. As shown in Figures 6 and 7, the content written in each category of the case summary output from the trained model will differ depending on the content of the character data input to the trained model.

[0055] An example of a prompt for generating a "topic" according to this embodiment is described below. In this embodiment, the "topic" is an item related to a change in a legal document. The prompt for generating a "topic" is an example of a second prompt according to this embodiment. The prompt may be disclosed to user P, or may be used only via the system without being disclosed.

[0056] As shown in Figure 8, the prompt for generating a "topic" in this embodiment preferably includes multiple steps. Each of the first to fourth steps in this embodiment includes a command statement for the trained model and an output example. By processing character data through multiple steps and with output examples attached to a large-scale language model in this way, the large-scale language model can appropriately output "topic" information.

[0057] The first and second steps are, for example, steps for gradually extracting matters related to changes in legal documents from input character data. In the first step, for example, topics under discussion are extracted from the character data, without being limited to matters related to legal documents. Then, in the second step, only topics related to changes in legal documents are extracted from the results extracted in the first step. By gradually extracting from the input character data in this manner, matters related to changes in legal documents can be extracted with high accuracy. More specifically, by including the first and second steps in the prompt for generating a "topic," extraction of topics other than matters related to changes in legal documents is reduced, and matters related to changes in legal documents are reduced from being overlooked.

[0058] Specifically, the prompt in the first step may include a command statement such as "Please briefly list the topics that were discussed in the email exchanges," and output examples of the topics that were discussed. In this embodiment, for example, as the output result of the first step, examples of the topics that were discussed include "Purpose of adjusting the product price," "Amendments to Articles n and m," "Matters to be approved," and "Related files."

[0059] The prompt in the second step may include a command such as, "Please select a topic that has had legal discussions about the topic obtained in step 1 and has had changes applied to the legal document," along with an output of the topic to which changes have been applied to the legal document. The output examples in the second step may be more specific than the examples in the first step. In this embodiment, the output in the second step may include "Product Price Modification" and "Amendments to Articles 3 and 5." While the first step did not specify article numbers, such as "Article n, Article m," the second step outputs specific article numbers.

[0060] The third and fourth steps extract arguments and their conclusions related to the topics extracted in the first and second steps. More specifically, the third step extracts arguments related to the topics extracted in the second step from the input character data. The fourth step extracts only the final conclusion from the results of the third step.

[0061] Specifically, the third step includes, for example, a command statement such as "Please explain, using expressions from the text, what kind of discussion took place regarding the topic in step 2." and an output example of a discussion related to the topic extracted in the second step. In this embodiment, in the third step, for the topic "Product Price Adjustment," a sentence such as "According to the client, 'We would like to change the price from 10 yen to 11 yen,' and the legal department responded, 'We understand.'" is output. Also, for the topic "Amendments to Articles 3 and 5," a sentence such as "According to the client, the other party requested that 'we ask you to revise Articles 3 and 5.' In response, the legal department responded, 'It is difficult to accept the other party's request,' and the client responded, 'We will share this information with the other party.'" is output.

[0062] The fourth step includes a command statement such as "Please extract a final conclusion from the output of Step 3," and an output example of a final conclusion corresponding to the topic extracted in the second step. In this embodiment, for the topic "Product Price Adjustment," the sentence "The legal department approved the change from 10 yen per item to 11 yen per item" is output. Also, for the topic "Amendments to Articles 3 and 5," the sentence "The legal department responded that it would be difficult to make adjustments to Articles 3 and 5" is output. In this way, by generating explanatory text for the discussion related to each topic in stages in the third and fourth steps, the accuracy of the output project summary is improved.

[0063] In the example shown in Fig. 8, the prompt for generating a "topic" has a text data input field, similar to the prompt for generating a "summary" shown in Fig. 5. Note that in Figs. 5 and 8, the prompt for generating a "summary" and the prompt for generating a "topic" each have a text data input field, but the prompt for generating a "summary" and the prompt for generating a "topic" may share a single text data input field.

[0064] The steps included in the prompt for generating a "topic" are not limited to the above examples. For example, the prompt for generating a "topic" may include additional steps. It is preferable that the prompt for generating a "topic" includes at least the first and second steps. This allows for the appropriate generation of a case summary using a large-scale language model.

[0065] Furthermore, the examples included in each step are merely examples and are not limited to those described above. The content of the examples may be customized to be different for each of users P1 to PN. Furthermore, when text other than the body of an email is input as character data, multiple phrases may be prepared to be written before the input of the character data, depending on the format of the text data. Furthermore, as described in the explanation of FIG. 5 , the description indicating the type of text data in the prompt may be changeable by user P. If the prompt is not disclosed to user P, for example, a screen on the browser of terminal device T1, in which the "type of text data" and the "text data" can be input, may be displayed by the display control unit 114 (described later) on the screen.

[0066] FIG. 9 is an example of a "topic" generated by the prompt in FIG. 8 . The case summary generation unit 113 inputs character data and the prompt for generating a "topic" shown in FIG. 8 into the trained model, causing the trained model to output the "topic" shown in FIG. 9 . As shown in FIG. 9 , the "topic" output from the trained model of this embodiment includes the title of an item related to a change in the legal document and a description of its content. The title of the item related to the change in the legal document is, for example, the topic extracted in the second step described above. The description of the content of the item related to the change in the legal document is the conclusion generated in the fourth step described above. In the example shown in FIG. 9 , the title of the item related to the change in the legal document is "Price Error," and the description of the item is "The client reported via email that 'File Name Assignment 13 yen (excluding tax) per item' needs to be corrected to 'File Name Assignment 54 yen (excluding tax) per item.'"

[0067] Figure 10 shows another example of a "topic" generated by the prompt in Figure 8. Figure 10 shows a "topic" output from a trained model to which character data different from the example shown in Figure 9 has been input. As shown in Figures 9 and 10, the content written in each category of the case summary output from the trained model will differ depending on the content of the character data input to the trained model.

[0068] Returning to FIG. 4, the display control unit 114 communicates information with the terminal device T1 via the network N, for example, to display various screens on the display device 30 of the terminal device T1.

[0069] Specifically, the display control unit 114 of this embodiment generates display information for displaying a legal case summary and outputs it to the terminal device T1. The display information may include layout information. In this embodiment, the display control unit 114 outputs to the terminal device T1 layout information that defines a screen layout for displaying the legal case summary in a browser, as well as text data and image data to be displayed on the screen. The layout information and data format transmitted by the display control unit 114 to the terminal device T1 may be executable by a general web browser, such as HTML (Hyper Text Markup Language) data or script (e.g., Java (registered trademark) Script). The display control unit 114 may display various screens on the display device 30 of the terminal device T1 by installing a dedicated application on the terminal device T1, for example. The display control unit 114 may also be referred to as a screen generation unit.

[0070] Here, examples of various screens displayed on the display device 30 of the terminal device T1 by the display control unit 114 will be described with reference to FIGS.

[0071] FIG. 11 is a diagram showing an example of a case display screen 300 according to this embodiment. The case display screen 300 according to this embodiment is a screen on which information related to a legal case is displayed. The display control unit 114 generates, for example, one case display screen 300 for each legal case. The example shown in FIG. 11 is a case display screen 300 on which information associated with the case number "2023-07-0001" is displayed.

[0072] The case display screen 300 includes, for example, a character data display area 320 and a tab display area 321. The display control unit 114 displays character data related to the legal case in the character data display area 320. In FIG. 11 , the text of an email is displayed in the character data display area 320 as an example. The case display screen 300 may also include multiple tabs 310a-310c. Hereinafter, when no particular distinction is made between the individual tabs 310a-310c, they will simply be referred to as tabs 310. The display control unit 114 displays, for example, the contents of one of the tabs 310a-310c selected by user P1 in the tab display area 321. In the example shown in FIG. 11 , the contents of the "Information" tab 310a are displayed in the tab display area 321. The "Information" tab 310a is a tab that displays case information, which is various information related to the legal case, such as the status of the legal case, the requesting department, the requester, the person in charge, and the deadline for a response. It should be noted that this information may be associated with, for example, case information and stored in the auxiliary storage device 15. The tabs 310a to 310c are an example of a plurality of screen areas in this embodiment.

[0073] Furthermore, the display control unit 114 of this embodiment displays a case summary generation button 301 on the case display screen 300. The case summary generation button 301 is a button image that can accept an instruction to generate a case summary from a user. The case summary generation button 301 is an example of a reception unit in this embodiment. When user P1 presses the case summary generation button 301, that is, when the case summary generation button 301 accepts an instruction to generate a case summary from user P1, the above-mentioned case summary generation unit 113 inputs character data, a prompt for generating a "summary," and a prompt for generating a "topic" into the trained model, and generates a "summary" and a "topic." Note that in this specification, a button is an example of a UI that a user operates to give an arbitrary instruction. An example of such a UI is a drop-down list.

[0074] The display control unit 114 can output the "summary" and "topic" generated by the case summary generation unit 113. In this embodiment, the display control unit 114 displays the generated "summary" and "topic" in an editable manner.

[0075] 12 is a diagram showing an example of a case summary editing screen 330 according to this embodiment. As shown in Fig. 12, the display control unit 114 according to this embodiment displays the case summary editing screen 330 as a pop-up on the case display screen 300. The case summary editing screen 330 may include a "summary" editing field 331, a "topic" editing field 332, and a registration button 333.

[0076] In this embodiment, the "summary" editing column 331 and the "topic" editing column 332 are screen areas in which the "summary" and "topic" generated by the case summary generation unit 113 are displayed in an editable manner. User P1 can check the "summary" and "topic" displayed in the "summary" editing column 331 and the "topic" editing column 332, and edit the text as necessary.

[0077] The registration button 333 in this embodiment is a button image that can accept an operation by user P1 to confirm and save the "summary" and "topic" displayed in the "summary" editing column 331 and the "topic" editing column 332.

[0078] When the registration button 333 is pressed, the registration processing unit 115 of the information processing device 10 of this embodiment stores the "summary" and "topic" displayed in the "summary" editing field 331 and the "topic" editing field 332 in, for example, the auxiliary storage device 15 of the information processing device 10. The display control unit 114 can display the confirmed "summary" and "topic" as a case outline of the legal case.

[0079] 13 is a diagram showing an example of a display mode of a case summary of a legal case according to this embodiment. In the example of FIG. 13, when the "Information" tab 310a is selected by user P1, a case summary display area 340 is provided in the tab display area 321. Furthermore, when the "Information" tab 310a is selected by user P1, the display control unit 114 displays a "Summary" and a "Topic" in the case summary display area 340 in the tab display area 321 of the case display screen 300.

[0080] The case summary display area 340 may be provided with an edit case summary button 341. The edit case summary button 341 is, for example, a button image that can receive an operation from the user P1 to instruct the user to start editing the "summary" and "topic." When the edit case summary button 341 is pressed by the user P1, the display control unit 114, for example, displays a pop-up case summary editing screen 330.

[0081] The case summary editing screen 330 may be in the pop-up format shown in FIG. 12, or may be in a case summary display area 340 within the tab display area 321 where the "summary" and "topic" can be edited.

[0082] 11 to 13 are merely examples, and the screen layout and various types of information displayed are not limited to the above examples. Also, only either the "Summary" or the "Topic" of the case overview may be displayed.

[0083] Furthermore, the screen displayed by the display control unit 114 is not limited to the above example. For example, the display control unit 114 may cause the display device 30 to display an input screen for the user P1 to input document data for generating a case summary of a legal case.

[0084] The information processing device 10 may also have other functions. For example, the information processing device 10 may have the document creation support function and the document viewing function described above. When the information processing device 10 has the document viewing function, for example, the user P1 may be able to perform a document review process on a browser screen such as the case display screen 300.

[0085] (Functional Configuration of Terminal Device T1) Next, the functional configuration of the terminal device T1 will be described. Fig. 14 is a diagram showing an example of functional blocks of the terminal device T1 according to this embodiment. The terminal device T1 includes, for example, a terminal control unit 211 and a browser unit 212 as functions realized by the processor 21. The functions realized by the terminal control unit 211 and the browser unit 212 are each stored as a program in, for example, the main storage device 23 or the auxiliary storage device 25. The processor 21 can realize various functions (the terminal control unit 211 and the browser unit 212) by reading and executing the programs stored in the main storage device 23 or the auxiliary storage device 25.

[0086] The terminal control unit 211 starts up the browser unit 212 in response to an instruction from the user P1, for example.

[0087] The browser unit 212 is, for example, a general web browser, and executes processing in accordance with HTML data, scripts, etc. When the browser unit 212 is started, for example, it transmits an initial access URL to the information processing device 10 and accesses it.

[0088] The browser unit 212 of this embodiment displays data including text data and image data received from the information processing device 10 in a predetermined format on the display device 30 of the terminal device T1. In other words, the browser unit 212 causes the display device 30 to display a screen generated by the display control unit 114 of the information processing device 10. For example, the browser unit 212 causes the display device 30 to display a case display screen 300 and a case summary editing screen 330 illustrated in Figures 11 to 13 based on the data and layout information received from the information processing device 10. The browser unit 212 may also cause the display device 30 to display an input screen for document data for generating a case summary for a legal case.

[0089] Furthermore, the browser unit 212 transmits operations performed by the user P1 on the screen displayed on the display device 30 to the information processing device 10 via the network N. For example, when the user P1 operates the input device 31 to press one of the tabs 310a to 310c, the case summary generation button 301, or the case summary edit button 341, the browser unit 212 transmits information that the button has been pressed to the information processing device 10. Furthermore, when the user P1 operates the input device 31 to edit the "summary" or "topic" on the case summary edit screen 330 and then presses the register button 333, the browser unit 212 transmits the "summary" and "topic" displayed on the case summary edit screen 330 to the information processing device 10.

[0090] In addition, when user P1 inputs character data for generating a case summary of a legal case into the input screen displayed on the display device 30, the browser unit 212 transmits the input character data to the information processing device 10.

[0091] (Case Summary Generation Process for Legal Cases) Next, the flow of the case summary generation process for legal cases in this embodiment will be described. FIG. 15 is a flowchart showing an example of the flow of the case summary generation process for legal cases in this embodiment. As a premise for the processing of this flowchart, the terminal control unit 211 of the terminal device T1 has already started up the browser unit 212 in response to an instruction from the user P1, for example. Also, it is assumed that the display control unit 114 of the information processing device 10 and the browser unit 212 of the terminal device T1 are displaying a case display screen 300 on the display device 30.

[0092] First, when an instruction to generate a case summary is received from the user P1 (S1 "Yes"), the case summary generation unit 113 of the information processing device 10 executes a case summary generation process (S2).

[0093] More specifically, when user P1 presses the generate case summary button 301 displayed on the case display screen 300, the browser unit 212 of terminal device T1 notifies the information processing device 10 that the generate case summary button 301 has been pressed. Then, the case summary generation unit 113 inputs character data related to the legal case for which a case summary is to be generated, a prompt for generating a "summary," and a prompt for generating a "topic" into the trained model, thereby causing the trained model to output a "summary" and a "topic." Note that the character data input to the trained model may be acquired in advance by the acquisition unit 111 and stored in the auxiliary storage device 15, etc., or may be input by user P1 together with an instruction to generate a case summary from user P1.

[0094] Until an instruction to generate a case summary is received, the display control unit 114 continues to display the case display screen 300 and waits in a state in which an instruction to generate a case summary can be received (S1 "No").

[0095] Then, the display control unit 114 displays the case summary generated by the trained model on the case summary editing screen 330 (S3). More specifically, the display control unit 114 of the information processing device 10 transmits layout information of the case summary editing screen 330 and text data of the case summary generated by the trained model to the terminal device T1, and the browser unit 212 of the terminal device T1 displays the case summary editing screen 330 including a "summary" and a "topic" on the display device 30 based on the layout information transmitted from the information processing device 10 and the text data of the case summary generated by the trained model.

[0096] Here, if the browser unit 212 of the terminal device T1 receives an operation to modify the "summary" or "topic" by the user P1 (S4 "Yes"), it changes the "summary" or "topic" on the case overview editing screen 330 to reflect the modification (S5).

[0097] Then, when the browser unit 212 of the terminal device T1 receives a confirmation operation by the user P1, for example, an operation of pressing the registration button 333 (S6 "Yes"), it transmits the "summary" and "topic" displayed on the case summary editing screen 330 to the information processing device 10. In this case, the registration processing unit 115 of the information processing device 10 stores the edited "summary" and "topic" in, for example, the auxiliary storage device 15 of the information processing device 10 (S7).

[0098] If the user P1 does not perform a confirmation operation (S6 "No"), the display control unit 114 of the information processing device 10 and the browser unit 212 of the terminal device T1 continue to display the case summary editing screen 330 in a state in which editing by the user P1 is possible. Furthermore, because it is not essential for the user P1 to modify the "Summary" and "Topic", the user P1 may perform a confirmation operation without accepting a modification operation by the user P1 (S4 "No").

[0099] When the "summary" and "topic" are saved based on the confirmation operation by user P1, the processing of this flowchart ends.

[0100] In this manner, the information processing device 10 of this embodiment inputs character data related to a legal case into a trained model, and outputs a case summary generated by the trained model based on the character data, which includes at least the history of the legal case and matters related to the review of legal documents related to the legal case. Therefore, the information processing device 10 of this embodiment can help the user P1, who is in charge of the legal case, and other users who will later become involved in the legal case, easily understand the case summary, including the history of the legal case. In other words, the information processing method executed by the information processing device 10 of this embodiment can provide a support technology that can improve the efficiency of using information related to legal work.

[0101] Furthermore, the information processing device 10 of this embodiment outputs the case summary separately from the "summary" and "topic." Therefore, the information processing device 10 of this embodiment allows the user P1 and other users to quickly grasp the flow of the past history of the legal case and matters related to the review of the legal document.

[0102] Furthermore, the information processing device 10 of this embodiment displays a case summary generation button 301 that can receive an instruction from user P1 to generate a case summary. When the case summary generation button 301 receives an instruction from user P1 to generate a case summary, character data related to the target legal case is input into the trained model and at least the sentences for the "summary" and "topic" are output. Therefore, the information processing device 10 of this embodiment allows user P1 to check the case summary of the legal case at a desired timing. The timing when user P1 issues an instruction to generate a case summary may be, for example, when the legal case is completed or when user P1's response has been completed and user P1 is waiting for the other party's response regarding the legal case. However, the timing is not limited to these examples, and user P1 may operate to generate a case summary at any timing.

[0103] Furthermore, the information processing device 10 of this embodiment displays information related to a legal case in multiple tabs 310a to 310c, and displays a case summary related to the legal case in one of the multiple tabs. Therefore, with the information processing device 10 of this embodiment, the user P1 can easily display or hide the case summary by selecting the tab 310a to 310c to be displayed.

[0104] Furthermore, the information processing device 10 of this embodiment inputs character data related to a legal case for which a case summary is to be generated and a prompt for generating a "summary" into the trained model. The prompt for generating a "summary" of this embodiment includes multiple categories included in the background and at least one example for each of the multiple categories. In response to this prompt, the information processing device 10 of this embodiment outputs a sentence regarding the background of the legal case, which includes multiple categories, generated by the trained model. Therefore, the information processing device 10 of this embodiment can provide user P1 with the history of the legal case organized by category.

[0105] Furthermore, the information processing device 10 of this embodiment inputs character data related to a legal case and a prompt for generating a "topic" into the trained model. The prompt for generating a "topic" includes at least a first step of extracting a topic that has been discussed from the character data and a second step of extracting matters related to changes in the legal document from the extraction result of the first step. Therefore, the information processing device 10 of this embodiment can accurately extract matters related to changes in the legal document from the input character data.

[0106] (Second embodiment) In the first embodiment described above, the information processing device 10 generated a case summary of a legal case in response to instructions from user P, but in this second embodiment, the information processing device 10 automatically generates a case summary regardless of instructions from the user.

[0107] The configuration of the information processing system S is the same as that of the first embodiment described with reference to Fig. 1. The hardware configuration and functional configuration of the information processing device 10 and the terminal device T1 are the same as those of the first embodiment described with reference to Figs. 2 to 4.

[0108] In addition to the same functions as those in the first embodiment, the case summary generator 113 of this embodiment generates a case summary when a legal case is created, for example, when the legal case is registered in the information processing device 100 and a case number is issued.

[0109] Furthermore, the case summary generation unit 113 of this embodiment can generate a case summary for a legal case each time progress is made in the legal case. The case summary generation unit 113 overwrites the previous case summary each time a new case summary is generated. A case in which progress is made in the legal case refers to, for example, when information related to the legal case is updated. A case in which information related to the legal case is updated includes a case in which new information is registered and a case in which existing information is changed. Specific examples of registering new information related to a legal case include updating a timeline related to the legal case or adding text data or files, such as sending or receiving emails. Updating a timeline or sending or receiving emails is also referred to as updating communication related to the legal case. Specific examples of changes to existing information may include, for example, changes to the legal case status, requesting department, client, responsible person, deadline, etc., included in the case information. Note that the previous case summary may not be overwritten and may be saved and viewable as an older version of the case summary.

[0110] In this embodiment, the case summary automatically generated by the case summary generator 113 can be modified by user P1. Furthermore, if user P1 modifies the automatically generated case summary, the case summary generator 113 stops automatic updating of the case summary for the legal case. In this case, the case summary generator 113 may be configured not to change or overwrite the case summary for the legal case unless instructed by user P1. In this case, even if communication related to the legal case is newly updated, the case summary generator 113 does not automatically overwrite the case summary modified by user P1. To stop automatic updating of the case summary, for example, a flag defining whether automatic updating is on or off may be provided in the auxiliary storage device 15. In this embodiment, the information processing device 100 stops automatic updating of the case summary modified by user P1, thereby preventing the text of the case summary edited by user P1 from being erased or rewritten by automatic updating. Note that the setting not to change or overwrite the case summary for the legal case may be set by the system. For example, when information from a threaded email is automatically imported and used as text data, a topic that is less relevant to the legal case may be started in the same email thread (e.g., a new legal case that is being discussed on a different point). In such a case, if a case summary is automatically generated, modified, or overwritten using information that is less relevant to the previous legal case, the legal case summary may not be appropriate. Therefore, if it is determined that no changes to the case summary are necessary based on the results of an analysis of the added text data and information differences, the case status, and other information, the system can be configured not to change or overwrite the case summary.

[0111] Furthermore, similar to the first embodiment, the case summary generation unit 113 of this embodiment can also generate a case summary when receiving an instruction to generate a case summary from user P. Even if user P makes modifications when initially generating a case summary, the case summary generation unit 113 can generate a new case summary and overwrite the previously generated case summary when receiving an instruction to generate a case summary from user P. For example, when user P presses the generate case summary button 301, the case summary generation unit 113 generates a new case summary and overwrites the previously generated case summary.

[0112] FIG. 16 is a flowchart showing an example of the processing flow for automatically generating a case summary of a legal case according to this embodiment.

[0113] If a new legal case is generated (S101 "Yes"), the case summary generation unit 113 of the information processing device 10 executes the case summary generation process (S102). On the other hand, if a legal case is not generated (S101 "No"), the case summary generation process is not triggered, and the process does not start.

[0114] Then, the registration processing unit 115 stores the case summary including the "summary" and "topic" generated by the case summary generation unit 113, for example, in the auxiliary storage device 15 of the information processing device 10 (S103). The stored case summary is displayed on the case display screen 300 by the display control unit 114, for example, the next time the case display screen 300 is opened by the user P. At this point, the processing of this flowchart ends.

[0115] 16 can be modified by the user P. Fig. 17 is a flowchart showing an example of the flow of a process for modifying a legal case summary according to this embodiment.

[0116] For example, the case display screen 300 of this embodiment has the same configuration as that of the first embodiment, and further includes an edit case summary button, which is a button image that can receive an instruction from user P1 to start revising the case summary. The create case summary button 301 may also function as the edit case summary button. When user P1 presses the edit case summary button, i.e., when user P1's instruction to start revising the case summary is received ("Yes" in S201), the display control unit 114 of this embodiment displays the generated case summary on the edit case summary screen 330 in an editable state (S202).

[0117] Furthermore, if an instruction to start revising the case summary by user P1 is not received (S201 "No"), the display control unit 114 continues to display the case display screen 300 and waits in a state where it can receive an instruction to start revising the case summary.

[0118] The browser unit 212 of the terminal device T1 then accepts the operation to modify the case summary by the user P1 (S203), and changes the case summary on the case summary editing screen 330 to reflect the modification (S204).

[0119] Then, when the browser unit 212 of the terminal device T1 receives a confirmation operation by the user P1, for example, an operation of pressing the register button 333 ("Yes" in S205), it transmits the "summary" and "topic" displayed on the case summary editing screen 330 to the information processing device 10. In this case, the case summary generation unit 113 of the information processing device 10 stops automatic updating of the case summary of the legal case that has been modified by the user P1 (S206). From this point on, in this case, the case summary generation unit 113 will not change or overwrite the case summary of the legal case unless instructed by the user P.

[0120] The registration processing unit 115 of the information processing device 10 stores the revised case summary in, for example, the auxiliary storage device 15 of the information processing device 10 (S207).

[0121] If the confirmation operation by the user P1 has not been received (S205 "No"), the display control unit 114 continues to display the case summary editing screen 330 and continues to be in a state in which the correction operation by the user P1 can be received. At this point, the processing of this flowchart ends.

[0122] In addition to manual updating by the user P1, there are cases where the case summary is automatically updated by the information processing device 10. Fig. 18 is a flowchart showing an example of the processing flow for automatically updating the case summary of a legal case according to this embodiment.

[0123] If information related to a legal case has been updated ("Yes" in S301) and automatic updates for the legal case have not been stopped ("No" in S302), the case summary generation unit 113 of the information processing device 10 executes a case summary generation process (S303). At this time, the case summary generation unit 113 generates a case summary by inputting the latest character data at the time of execution of the case summary generation process into the trained model. That is, the case summary generation unit 113 generates a new case summary by taking into account the information updated in S301.

[0124] Then, the registration processing unit 115 overwrites and saves the case summary including the "summary" and "topic" generated by the case summary generation unit 113 onto an existing case summary stored, for example, in the auxiliary storage device 15 of the information processing device 10 (S304).

[0125] If the information related to the legal case has not been updated (S301 "No"), the case summary generation process is not triggered and the process does not start. Also, even if the information related to the legal case has been updated, if automatic updates for the legal case have been stopped (S302 "Yes"), the case summary generation process is not executed and the process of this flowchart ends.

[0126] In this manner, the information processing device 100 of this embodiment generates a case summary related to a legal case when a new legal case is created, and overwrites and updates the generated case summary when information related to the legal case is updated. Therefore, the information processing device 100 of this embodiment can generate a case summary without requiring instructions from user P1 when a new legal case occurs or progress is made in a legal case. This reduces the effort required for user P1 to generate or update a case summary every time a legal case occurs or progress is made. Furthermore, the information processing device 100 of this embodiment automatically updates the case summary when a new legal case occurs or progress is made in a legal case, thereby reducing the occurrence of discrepancies between the latest status of the legal case and the case summary displayed on the screen.

[0127] In this embodiment, the case summary generator 113 automatically updates the case summary both when a new legal case is created and when information related to the legal case is updated, but the automatic update may be performed only in one of the cases. That is, the case summary generator 113 may generate a case summary related to a legal case at least when a new legal case is created or when information related to the legal case is updated.

[0128] (Variant 1) In each of the above-described embodiments, the display control unit 114 displays a case summary including a "summary" and a "topic" in the case summary display area 340 within the tab display area 321, but the display manner of the case summary is not limited to this.

[0129] For example, the display control unit 114 of this modified example always displays the case summary in the header section at the top of the case display screen 300 .

[0130] With this display, when the user P1 opens the case display screen 300 in the information processing device 10 of this modified example, the user P1 can immediately check the case outline including the "summary" and "topic."

[0131] Alternatively, the display control unit 114 may provide a button image on the case display screen 300 that can switch between displaying and hiding the case summary, and switch between displaying and hiding the case summary when the button image is pressed by user P1.

[0132] (Variation 2) In the first embodiment described above, the case summary "Summary" and "Topic" are generated in response to an instruction from user P1. In the second embodiment described above, the case summary is generated when a legal case is created. However, the trigger for generating the case summary is not limited to these.

[0133] For example, the case summary generation unit 113 of the information processing device 10 may automatically generate a case summary when at least one of the following occurs: when a legal case is completed; when user P1 is waiting for a response from the other party regarding the legal case; when information regarding the legal case has not been updated for a specified period of time; or when the legal case is associated with another legal case. The specified period is not particularly limited, but may be, for example, one week. Furthermore, user P1 or an administrator of the information processing device 10 may be able to set the specified period. Completion of a legal case occurs, for example, when the user changes the status field of the case information in the tab display area 321 to "Completed." Examples of when a legal case is associated with another legal case include when it is estimated that the legal case is similar to the other legal case and therefore should be referenced by the user; or when it is estimated that the legal case and the legal case will have an influence on each other, such as when a legal case is concluded in a master contract and an individual contract is concluded based on the master contract.

[0134] Furthermore, a case summary may be generated when a user other than the user in charge of the case accesses the case. This allows the legal case summary to be generated at a time that minimizes waste. This is particularly preferable because it can reduce costs when, for example, the trained model that generates the case summary exists outside the information processing system S and is different from the operator of the information processing system S or the user P, and when the billing method is a pay-per-use method for the generation of the case summary. Furthermore, the operation to generate the case summary may be performed by a user other than the user in charge of the case. The timing of the generation instruction is arbitrary, and can be performed, for example, when it is desired to refer to the response to the legal case as a similar case from the past.

[0135] For example, if the case summary generation unit 113 automatically generates a "summary" and a "topic," the display control unit 114 will display a case summary including the "summary" and the "topic" the next time user P1 opens the case display screen 300.

[0136] Furthermore, the case summary generator 113 of the information processing device 10 may execute the case summary generation process in response to a plurality of triggers. For example, the case summary generator 113 may generate a case summary of a legal case in any of the following cases: when the legal case is completed, when the user is waiting for a response from the other party regarding the legal case, and when information regarding the legal case has not been updated for a specified period of time.

[0137] (Variation 3) In the above-described embodiment, the trained model is stored outside the information processing device 10. However, the trained model may be stored, for example, in the auxiliary storage device 15 of the information processing device 10. Furthermore, the trained model may be incorporated into the case summary generation unit 113 of the information processing device 10.

[0138] The trained model may also be different for each company to which users P1 to PN belong. For example, the trained model may be a model trained using different training data according to the needs of each company. Alternatively, the trained model may be fine-tuned according to the needs of each user or the organization to which the user belongs. For the same purpose, a separate trained model may be provided that appropriately processes data before inputting or after outputting information to a common external trained model. Furthermore, while using a general-purpose trained model, additional training, fine-tuning, or a separate trained model may be provided to enable appropriate interpretation of legal cases, such as information on legal practice and laws and regulations, or to obtain appropriate output for summarizing cases.

[0139] In addition, in the above-described embodiment, it is assumed that user P1 can modify the "summary" and "topic" output from the trained model, but the trained model may also learn the results of modifications made by user P1 to improve the accuracy of generating project summaries.

[0140] (Variation 4) In the above-described embodiment, the case summary generation unit 113 generates "summaries" and "topics" from character data using a trained model. However, the case summary generation unit 113 may generate "summaries" and "topics" by further referring to information other than character data. For example, the case summary generation unit 113 may extract "related contracts" included in the "summary" and topics to which changes have been applied to legal documents included in the "topics" based on, for example, update logs of legal documents related to the legal case for which the case summary is to be generated. This extraction may be performed by a trained model or by another search function, etc.

[0141] Although the embodiments of the present disclosure have been described in detail above, these embodiments are presented as examples and are not intended to be limiting. Each embodiment can be modified by various additions, changes, substitutions, partial deletions, combinations, etc., without departing from the technical spirit of the present invention. These embodiments and their modifications are included in the scope of the invention described in the claims and their equivalents.

[0142] DESCRIPTION OF SYMBOLS 10 Information processing device 11 Processor 13 Main memory device 15 Auxiliary memory device 17 Network interface 18 Device interface 19 Bus 21 Processor 23 Main memory device 25 Auxiliary memory device 27 Network interface 28 Device interface 29 Bus 30 Display device 31 Input device 111 Acquisition unit 112 Operation reception unit 113 Case summary generation unit 114 Display control unit 115 Registration processing unit 211 Terminal control unit 212 Browser unit 300 Case display screen 301 Case summary generation button 310a to 310c Tabs 320 Character data display area 321 Tab display area 330 Case summary editing screen 331 Edit field for "Summary" 332 Edit field for "Topic" 333 Registration button 340 Case summary display area 341 Case summary editing button N Network P1 to PN User S Information processing system T1 to TN Terminal device

Claims

1. An information processing method comprising: inputting character data related to a legal case into a trained model; and outputting a case summary generated by the trained model based on the character data, the case summary relating to at least the history of the legal case and matters related to the review of legal documents related to the legal case.

2. The information processing method of claim 1, further comprising: the case summary including a first sentence relating to the history of the legal case and a second sentence relating to matters related to the review of the legal document; and outputting the first sentence and the second sentence separately.

3. The information processing method of claim 1, further comprising: generating the case summary related to the legal case when the legal case is newly created; and overwriting and updating the generated case summary when information related to the legal case is updated.

4. The information processing method of claim 1, further comprising: displaying a reception unit capable of receiving an instruction from a user to generate a case summary; when the reception unit receives an instruction from the user to generate a case summary, inputting the character data into the trained model; and outputting the case summary generated by the trained model.

5. The information processing method of claim 1, further comprising: inputting the character data related to the legal case into the trained model; and outputting the case summary generated by the trained model when at least one of the following occurs: the legal case is completed, the user is waiting for a response from the other party regarding the legal case, and information regarding the legal case has not been updated for a specified period of time.

6. The information processing method of claim 1, further comprising: displaying information related to the legal case in a plurality of screen areas; displaying the case summary related to the legal case in one of the plurality of screen areas; and displaying the legal document in one of the plurality of screen areas.

7. The information processing method of claim 1, further comprising: displaying a screen on which information relating to the legal case is displayed; and displaying the case summary relating to the legal case at the top of the screen.

8. The information processing method of claim 1, further comprising: inputting the character data related to the legal case and a first prompt into the trained model, the first prompt including a plurality of categories included in the past history and at least one example for each of the plurality of categories; and outputting a sentence regarding the past history of the legal case including the plurality of categories generated by the trained model.

9. An information processing method as described in any one of claims 1 to 8, further comprising inputting the character data related to the legal case and a second prompt into the trained model, the second prompt including at least a first step of extracting a topic under discussion from the character data, and a second step of extracting matters related to changes to the legal document from the extraction result of the first step.

10. An information processing device comprising at least one processor, the at least one processor inputting character data related to a legal case into a trained model, and outputting a case summary generated by the trained model based on the character data, the case summary relating to at least the history of the legal case and matters related to the review of legal documents related to the legal case.

Images