Document correction support program, document correction support system, and document correction support method
The document correction assistance program addresses the issue of inaccurate revision proposals by using a generation device to create tailored correction proposals based on input and related document information, resulting in more accurate and appropriate document revisions.
Patent Information
- Application Number
- JP2023192925
- Authority / Receiving Office
- JP · JP
- Patent Type
- Applications
- Current Assignee / Owner
- Filing Date
- 2023-11-13
- Publication Date
- 2025-05-23
AI Technical Summary
Existing document correction assistance technologies, such as those described in Patent Document 1, fail to generate revision proposals that accurately align with the content of the input document, often resulting in inappropriate corrections.
A document correction assistance program that acquires input document information and related document information, uses a generation device to generate correction proposals based on the input information, and transmits these proposals for use in revising the input document.
The solution enables the generation of revision proposals that are tailored to the specific content of the input document, improving the accuracy and appropriateness of document corrections.
Smart Images

Figure 2025079989000001_ABST
Abstract
Description
[Technical field]
[0001] The present invention relates to a document correction assistance program, a document correction assistance system, and a document correction assistance method. [Background technology]
[0002] Generally, when a company or other entity concludes a contract, the legal staff must review the contract and make repeated revisions to the content to make mutual adjustments. In such work, contracts with similar content are often concluded repeatedly, which tends to result in inefficient work of making the same revisions and comments each time. There was also the problem that even though the contracts had similar content, there was variation in the perspective and quality of the comments depending on the person in charge.
[0003] Regarding this issue, there is known a technique for supporting revisions based on the descriptions in other contracts, etc. For example, Patent Document 1 describes a technique for displaying the contents of similar contracts as a reference for revising an input contract. [Prior art documents] [Patent documents]
[0004] [Patent Document 1] International Publication No. 2020 / 230346 Summary of the Invention [Problem to be solved by the invention]
[0005] However, the technology described in Patent Document 1 merely displays the text after the corrections made to similar contracts as correction candidates based on the corrections made thereafter. Therefore, there is a problem that the corrections proposed are not necessarily in line with the contents of the input contract that is actually to be corrected. Therefore, the present invention aims to solve the above-mentioned problems. [Means for solving the problem]
[0006] In order to solve the above problems, the present invention provides a document correction assistance program for assisting in document correction, which causes a computer to function as an acquisition unit, a correction unit, and an output unit, wherein the acquisition unit acquires input document information indicating at least a portion of the content of an input document, which is a legal document to be corrected, and related document information indicating at least a portion of the content of a related document, which is a legal document related to the input document, the correction unit inputs the input document information and the related document information to a generation device, causes the generation device to generate a correction proposal for the input document information, acquires the correction proposal, and the output unit transmits the correction proposal.
[0007] With this configuration, it is possible to generate a revision proposal based on the contents of the input document, thereby solving the above-mentioned problem and generating an appropriate revision proposal according to the actually input document.
[0008] In a preferred embodiment of the present invention, the related document information further includes information regarding corrections made in creating the related document, and the correction unit further inputs information regarding the corrections of the related document to the generation device to generate the correction proposal.
[0009] With this configuration, the proposed amendment can be generated based on past amendments made to the related document, thereby generating a more appropriate proposed amendment taking into account the contents of the input document and the contents of amendments made to the past legal document.
[0010] In a preferred embodiment of the present invention, the information regarding the revision includes a comment indicating the content of the revision in the related document.
[0011] With this configuration, it is possible to generate a revision proposal based on comments added to related documents, thereby generating a more appropriate revision proposal by taking into account the content of the input document and the content of comments added to past legal documents.
[0012] In a preferred embodiment of the present invention, the information regarding the revision includes information indicating the contents of at least a portion of the related document before and after the revision.
[0013] With this configuration, it is possible to take into consideration the portion that has been corrected and the viewpoint of the correction when generating a correction proposal, and therefore it is possible to generate a more appropriate correction proposal.
[0014] In a preferred embodiment of the present invention, the information on the revision further includes information on a revision location identified based on a difference between contents of the related document before and after the revision.
[0015] In a preferred embodiment of the present invention, the information on the revision includes information on the revision location in the related document.
[0016] With this configuration, it is possible to generate a revision proposal taking into account parts that have been revised in the past. This makes it possible to prevent problems such as erroneously correcting differences in wording between individual contracts that do not require revision, and to generate a more appropriate revision proposal.
[0017] In a preferred form of the present invention, the information regarding the corrections includes, for each correction, information about the user who made the correction, and the correction unit causes the generation device to create the correction proposal based on information about corrections to the related document made by a user in the same position as the user who submitted the input document information.
[0018] With this configuration, it is possible to generate a revision proposal based on revisions made by users in the same position, such as other employees of the same company, legal counsel, etc. This makes it possible to generate a more appropriate revision proposal that takes into account revisions that reflect the company's desired conditions.
[0019] In a preferred form of the present invention, the generation device is a device that generates the revision suggestion using a sentence generation model, and the correction unit generates revision instructions in natural language based on the input document information and related document information and transmits them to the generation device, thereby generating the revision suggestion.
[0020] With this configuration, it becomes possible to generate revision suggestions using an interactive text generation model.
[0021] In a preferred form of the present invention, the input document information is information indicating the contents of one of the correction units when the input document is divided into predetermined correction units, the acquisition unit acquires a plurality of pieces of input document information for one of the input documents, and the correction unit causes the generation device to generate the correction proposal for each piece of input document information.
[0022] With this configuration, a revision proposal can be generated for each group of revisions.
[0023] In a preferred form of the invention, the input document and related documents are legal documents containing a plurality of clauses, and the amendment unit is a clause or a group of clauses defining a set of provisions contained in the input document or related documents.
[0024] With this configuration, amendment proposals can be generated for each clause or group of clauses, which allows the user to consider amendments at the unit of provisions defined by the legal document, improving convenience.
[0025] In a preferred form of the present invention, the computer further functions as an evaluation unit, which acquires evaluation information including suggestions associated with a portion of the input document, and the correction unit is the portion of the input document that corresponds to the suggestions.
[0026] With this configuration, it is possible to show a correction proposal for each of the findings that indicate the evaluation results for the input document. This makes it possible to show the evaluation results for the input document and the correction proposal together, thereby more effectively supporting document correction.
[0027] In a preferred form of the present invention, the computer is further made to function as an identification unit, the acquisition unit further acquires attributes of the input document, and the identification unit identifies documents that have commonality with the input document as the related documents based on the attributes of the input document from a database in which a plurality of documents and their attributes are registered.
[0028] In a preferred form of the present invention, the input document is a legal document relating to an agreement with a specific party, and the attributes of the input document and the attributes of the document registered in the database include any of the party to the agreement, the user's position in the agreement, the type of the agreement, and the subject of the agreement.
[0029] With this configuration, for example, documents with similar contents can be selected as related documents based on the contents of the agreement specified by the input document, conditions related to the agreement, etc. This makes it possible to generate a more appropriate revision proposal.
[0030] In a preferred embodiment of the present invention, the identification unit identifies a plurality of the related documents, and the correction unit transmits the related document information based on a user's selection from among the identified plurality of related documents to the generation device, thereby generating the correction proposal.
[0031] With this configuration, the related documents used to generate a revision proposal can be selected in a manner that reflects the user's intentions, and this is expected to result in the generation of a more appropriate revision proposal.
[0032] In a preferred form of the present invention, the database stores legal documents that have been created in the past during the process of an agreement as the documents, and for each of the documents, stores a status tag indicating the stage in the adjustment of the agreement that each document specifies, and the identification unit identifies the related documents based on the status tag.
[0033] With this configuration, amendment proposals can be generated based on legal documents from actual past agreements. Also, it becomes possible to identify related documents depending on the stage of the agreement, which is expected to have the effect of generating more appropriate amendment proposals.
[0034] In a preferred form of the present invention, the identification unit identifies multiple documents related to the same agreement that have different status tags assigned to them as the related documents, and the correction unit transmits information regarding corrections to the related documents generated based on the differences between the multiple documents to the generation device to generate the correction proposal.
[0035] With this configuration, information regarding revisions can be automatically obtained from the differences in past documents. As a result, even if data including information regarding revisions is not available, it is possible to generate a revision proposal for an input document based on revisions to past documents.
[0036] In a preferred form of the present invention, the identification unit prioritizes identifying as the related document a document that has been assigned a presentation tag indicating that the document has been presented to the other party in the agreement by a party in the same position as the user who sent the input document during the adjustment of the agreement.
[0037] Generally, a document that has been presented to a third party strongly reflects the intentions of the company. Therefore, by adopting this configuration, it is expected that a revision proposal that reflects the intentions of the company can be generated.
[0038] In a preferred embodiment of the present invention, when there are multiple documents to which the presentation tag is assigned, the identification unit prioritizes identifying as the related document the document that was created at an earlier stage in the agreement specified by each document.
[0039] With this configuration, it is possible to generate an amendment proposal based on documents presented at an earlier stage. When making an agreement, it is common to first sound out the other party with a proposal that strongly reflects one's own intentions, and then make adjustments based on the other party's comments and amendments. Therefore, by creating an amendment proposal based on documents presented by one's own side at an earlier stage, it is expected that an amendment proposal with terms more suitable for one's own company will be created.
[0040] In a preferred embodiment of the present invention, the specification section specifies a plurality of the related documents and determines priorities of the plurality of the related documents such that a document to which the presentation tag is added has a higher priority.
[0041] In a preferred form of the present invention, when there are multiple documents to which the presentation tag is assigned, the identification unit determines the priority of the multiple related documents so that the document created at an earlier stage in the agreement specified by each document is given a higher priority.
[0042] In a preferred embodiment of the present invention, the revision section inputs the related document information relating to the related document with the highest priority among the plurality of related documents to the generation device, thereby generating the proposed revision.
[0043] With this configuration, revision proposals can be generated using related documents with higher priority, i.e., related documents that are considered to more strongly reflect the company's intentions, thereby generating more appropriate revision proposals.
[0044] In a preferred embodiment of the present invention, the computer further functions as a display unit, The display unit displays the related document information relating to the plurality of related documents in a priority order.
[0045] With this configuration, related documents with higher priority, i.e., related document information that is considered to more strongly reflect the intentions of the company, can be presented to the user and provided as reference information for revisions.
[0046] In a preferred form of the present invention, the database stores multiple legal documents that were created in the past during the course of the same agreement along with information indicating their chronological order as the documents, and among the documents related to the same agreement to which the presentation tag is assigned, documents with older chronological order are given priority in identifying them as the related documents.
[0047] In a preferred embodiment of the present invention, the correction section further obtains a comment indicating the content of the correction to the input document, and the output section further transmits the comment.
[0048] With this configuration, reference information regarding corrections can be provided.
[0049] In order to solve the above problems, the present invention provides a document correction support system for supporting document correction, comprising an acquisition unit, a correction unit, and an output unit, wherein the acquisition unit acquires input document information indicating at least a portion of the content of an input document, which is a legal document to be corrected, and related document information indicating at least a portion of the content of a related document, which is a legal document related to the input document, the correction unit inputs the input document information and the related document information to a generation device, causes the generation device to generate a correction proposal for the input document information, acquires the correction proposal, and the output unit transmits the correction proposal.
[0050] In order to solve the above problems, the present invention provides a document correction support method for supporting the correction of a document, wherein a computer acquires input document information indicating at least part of the content of an input document which is a legal document to be corrected, and related document information indicating at least part of the content of a related document which is a legal document related to the input document, inputs the input document information and the related document information into a generation device to cause the generation device to generate an amendment to the input document information, acquires the amendment, and transmits the amendment.
Effect of the Invention
[0051] According to the present invention, an amendment can be generated in accordance with the content of the input document for which correction is actually to be made.
Brief Description of the Drawings
[0052] [Figure 1] Block diagram showing the configuration of the system of the present embodiment. [Diagram 2] Hardware configuration diagram of the information processing apparatus and the terminal apparatus of the present embodiment. [Diagram 3] Block diagram showing the functional configuration of the system of the present embodiment. [Figure 4] Example of information related to the document of the present embodiment. [Diagram 5] Sequence diagram related to the amendment generation process of the present embodiment. [Figure 6] Example of screen display of the present embodiment. [Figure 7] Example of screen display of the present embodiment. [Figure 8] Example of screen display of the present embodiment. [Figure 9] Example of screen display of the present embodiment. [Figure 10] Flowchart related to the related document specifying process of the present embodiment.
Mode for Carrying Out the Invention
[0053] The present invention will now be described in more detail with reference to the accompanying drawings, in which preferred embodiments are shown, but which may be embodied in many different forms and are not limited to the embodiments set forth herein.
[0054] For example, in this embodiment, the configuration, operation, etc. of the document revision support system will be described, but similar effects can be achieved by a device having similar functions, a method executed by the device, a computer program that causes a computer device to execute the method, etc. In this embodiment, the document revision support system takes the form of so-called cloud computing, in which a document revision support program is executed in a server (document revision support device) connected to the client terminal via a network in order to realize the functions in the client terminal (user terminal).
[0055] The document correction support program may be provided as a non-transient computer-readable recording medium, or may be provided so as to be downloadable from an external server. The document correction support program may be, for example, a program that is added on (add-in, plug-in) to document editing software such as Microsoft Word (registered trademark).
[0056] In the following, the term "unit" may include, for example, a combination of hardware resources implemented by a circuit in the broad sense and software information processing that can be specifically realized by these hardware resources. In this embodiment, "information" is represented by, for example, a physical value of a signal value representing a voltage or current, a high or low signal value as a binary bit collection consisting of 0 or 1, or a quantum superposition (so-called quantum bit), and communication and calculation can be performed on a circuit in the broad sense.
[0057] A circuit in the broad sense is a circuit realized by appropriately combining a circuit, a processor, a memory, etc. For example, it is a circuit including any of a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), an LSI (Large Scale Integration), an ASIC (Application Specific Integrated Circuit), an FPGA (Field-Programmable Gate Array), etc.
[0058] <1.Definition> First, definitions of terms will be explained. The present invention relates to a technique for supporting document revision. A document to be revised and input to the system of the present invention is called an input document. Furthermore, a document related to the input document and used as a reference for the revision of the input document is called a related document. In the present invention, both the input document and the related document are legal documents.
[0059] Legal documents are documents that are somehow related to the law, and include contracts, rules, regulations, notices, agreements, notices, powers of attorney, etc. Possible rules include documents of legal norms such as work regulations. Possible notices include documents declaring policies such as privacy. In this embodiment, it is assumed that the input document and related documents are legal documents, particularly contracts, relating to arrangements with specific parties.
[0060] Here, "agreement" refers to a decision made between two parties, including terms of use, declarations to users, agreements with other parties, etc. Examples of contracts include non-disclosure agreements, sales contracts, real estate lease agreements, and other documents that have contractual effect. Some legal documents include units divided into articles, paragraphs, subsections, etc., and in this embodiment, each of such units is called a "clause." An example of an embodiment of the present invention will be described in detail below.
[0061] <2. System configuration> Fig. 1 is a block diagram showing the configuration of a document correction support system according to this embodiment. As shown in Fig. 1, the document correction support system 0 includes a document correction support device 1 and a user terminal 2. The document correction support device 1 and the user terminal 2 are configured to be able to communicate with each other via a network NW. The document correction support device 1 operates as a server, and the user terminal 2 operates as a client terminal. In this embodiment, the network NW is an IP (Internet Protocol) network, but there are no limitations on the type of communication protocol, the type of network, and the like.
[0062] The document editing support device 1 is further connected to a generating device 3 via a network NW. The generating device 3 is an external server device that provides a text generation model. The document editing support device 1 may store the text generation model and may be configured to be able to edit a document without using an external device, or a part or all of the processing executed by the document editing support device 1 in this embodiment may be executed in the user terminal 2.
[0063] As the document editing support device 1, the generating device 3, and the database server, one or more information processing devices 10 (computer devices) such as a general-purpose server or a personal computer can be used. In this embodiment, the document editing support device 1 is a terminal device 90 having installed therein a computer program (document editing support program) that executes the document editing support method.
[0064] A terminal device 90 (computer device) such as a personal computer, a smartphone, or a tablet terminal can be used as the user terminal 2. The user terminal 2 has an application (web browser application or dedicated application) for making a request to the document editing support device 1 and receiving a response, and uses the processing results of the document editing support program executed in the document editing support device 1 via this application.
[0065] The document revision support device 1 transmits data in a predetermined format as various display information to the application of the user terminal 2 based on a request transmitted from the application of the user terminal 2. The data in the predetermined format is, for example, data in JSON (JavaScript (registered trademark) Object Notation) format or XML format. The application of the user terminal 2 renders display data based on the received data, and displays related document information, revision suggestions, and the like on the display of the user terminal 2.
[0066] For example, when a display process of various information is executed on a web browser application of the user terminal 2, the web browser application first transmits an HTTP (Hypertext Transfer Protocol) request to a web server and receives an HTTP response including the file requested by the HTTP request. The received file is display information including a revision proposal, indicates the contents or instructions of the web page to be displayed, and is described using a markup language such as HTML (Hypertext Markup Language). A script, which is a program executed on the web browser application, may be attached to the received file. The web browser application draws (renders) a web page in the display area based on the received file. If a script is attached to the file, the browser executes the script and displays the web page according to the execution result.
[0067] <3. Hardware configuration> Fig. 2(a) is a hardware configuration diagram of the information processing device 10. As shown in Fig. 2, the information processing device 10 has a control unit 101, a storage unit 102, and a communication unit 103, which are used to cause each unit and each process to function.
[0068] The control unit 101 has a processor such as a CPU capable of executing an instruction set, and executes an OS and programs. The storage unit 102 includes a volatile memory such as a RAM capable of storing an instruction set, and a non-volatile recording medium such as an HDD or SSD capable of recording an OS, an XX program, a DBMS (database server), and the like. The communication unit 103 has an interface for physically connecting to a network, and controls communication with the network NW to input and output information.
[0069] Fig. 2(b) is a hardware configuration diagram of the terminal device 90. As shown in Fig. 2, the terminal device 90 has a control unit 901, a storage unit 902, a communication unit 903, an input unit 904, and an output unit 905, which are used to perform the functions of each unit and each process.
[0070] The control unit 901 has a processor such as a CPU capable of executing an instruction set, and executes an OS, programs, and the like. The storage unit 902 includes a volatile memory such as a RAM capable of storing an instruction set, and a non-volatile recording medium such as an HDD or SSD capable of recording an OS, a determination program, and the like. The communication unit 903 has an interface for physically connecting to a network, and controls communications with the network NW to input and output information. The input unit 904 includes an operation input device capable of input processing, such as a touch panel or a keyboard, and an audio input device capable of audio input, such as a microphone. The output unit 905 includes a display device capable of display processing, such as a display, and an audio output device, such as a speaker.
[0071] <4. Functional configuration> 3 is a diagram showing the functional configuration of the document correction support device 1, the user terminal 2, and the generation device 3 in this embodiment. The document correction support device 1 has an acquisition unit 11, an evaluation unit 12, an identification unit 13, a correction unit 14, an output unit 15, and a database DB. This is a concrete realization of information processing by software using hardware.
[0072] The acquisition unit 11 acquires input document information indicating at least a part of the content of the input document and related document information indicating at least a part of the content of the related document. The acquisition unit 11 also acquires attributes of the input document together with the input document information. In this embodiment, the acquisition unit 11 acquires a part or the whole of the input document, and divides the acquired information into multiple units to acquire the input document information.
[0073] The input document information in this embodiment is information indicating the contents of a part of the input document that corresponds to one correction unit when the input document is divided into predetermined correction units. That is, the acquisition unit 11 in this embodiment divides the input information into correction units and acquires multiple pieces of input document information. The correction unit is, for example, each clause, or a group of multiple clauses that defines a set of regulations. In addition, when an evaluation is performed on the input document and corrections are made based on the findings associated with a part of the input document as a result of the evaluation, the part of the input document that corresponds to the findings may be used as the correction unit.
[0074] The related document information is information that indicates the contents of a part of the related document that corresponds to the input document information. For example, if the input document information is information that indicates the contents of a clause that defines a certain provision in the input document, the related document information is information that indicates the contents of a clause that defines a similar provision in the related document.
[0075] The related document information of this embodiment also includes information regarding corrections made to the portion indicated by the related document information when the related document was created. The information regarding the corrections may be, for example, information regarding the description before and after the correction, the correction location, the user who made the correction, comments regarding the correction, etc. The correction location may be identified based on the difference between the description before and after the correction, and is information that identifies a position in the document.
[0076] The evaluation unit 12 acquires evaluation information including findings associated with a part of the input document. The evaluation information may be acquired from an external device that evaluates the input document, but in this embodiment, the evaluation unit 12 acquires the evaluation information by executing an evaluation process on the input document.
[0077] The evaluation information includes, for each of one or more issues, a clause in the input document that corresponds to the issue and the issue itself. The issue itself is a sentence that suggests a correction or points out a problem with the contents of the input document, such as "Is it necessary to specify a consultation clause?" or "Isn't the period of confidentiality set longer than necessary?" In addition, the issue itself may include information such as an explanatory text about the clause related to the issue, a template for the clause, and a risk level indicating the importance of the issue. The issue is a legal evaluation, for example, an evaluation from a perspective that may become a legal issue.
[0078] Specifically, the acquisition unit 11 acquires at least a part of the input document, and the evaluation unit 12 acquires the information. Then, the input document is evaluated based on the attributes of the input document. The evaluation may be performed by any method, such as evaluation by a machine learning model or rule-based evaluation. For example, when the rule-based evaluation is applied, the evaluation unit 12 determines which of a plurality of predetermined findings (evaluation comments) should be associated with each part (e.g., clause) of the input document, based on keywords included in the part of the input document and a predetermined keyword criterion (rule-based model). When the machine learning evaluation is applied, the evaluation unit 12 determines the findings to be associated with the part of the input document by a machine learning model that outputs an estimation result that is a result of estimating which of a plurality of predetermined findings (evaluation comments) should be associated based on the contents of the part of the input document that has been input. Alternatively, a sentence generation model may be applied as a type of machine learning model, and the evaluation unit 12 may generate findings for a part of the input document under a predetermined evaluation rule. The contents of each clause and the relationships between multiple clauses can be evaluated using, for example, the contract type and contractual position as attributes, and the title of each clause in the input document. The evaluation can be performed from the perspective of whether unnecessary clauses (or clauses that are disadvantageous to the party in the relationship with the other party included in the attributes of the input document) are included, whether there are clauses that should be added, whether the contents of the clauses are appropriate, etc.
[0079] The identification unit 13 identifies related documents from among documents registered in the database DB based on the attributes of the input document acquired by the acquisition unit 11. The identification unit 13 of this embodiment identifies a plurality of related documents based on the commonality of attributes between the input document and documents registered in the database DB, and the status tags of the documents registered in the database DB. The process of identifying related documents will be described later.
[0080] The correction unit 14 inputs the input document information and the related document information to the generation device 3, causes the generation device 3 to generate a correction proposal corresponding to the input document information, and acquires the correction proposal. In this embodiment, an example will be described in which the correction unit 14 generates a correction command in natural language and transmits it to the generation device 3, and the generation device 3 generates the correction proposal using a sentence generation model. The sentence generation model refers to any model that can output a sentence. As the sentence generation model, a model generated by machine learning can be representatively used. In this embodiment, a case will be described in which a conversational sentence generation model is used, which receives a sentence and generates a sentence as a response to the input sentence.
[0081] More specifically, the sentence generation model is a machine learning model for natural language processing, and is a model that has a function of generating sentences based on input information. For example, the sentence generation model is a machine learning model based on a Transformer model having a self-attention mechanism. The Transformer model is, for example, a GPT (Generative Pre-trained Transformer) model. The sentence generation model can perform learning using a large data set, enabling not only general natural language generation tasks but also highly specialized sentence generation tasks. For example, the sentence generation model can be a model trained using a large data set using the GPT-3 or GPT-4 algorithm developed by OpenAI (for example, a ChatGPT model generated by combining learning using a large data set and reinforcement learning using a reward prediction model). In addition, in this embodiment, an example in which the input to the model is text information will be described, but image information or audio information may also be input.
[0082] The correction command includes input document information and related document information, and is information indicating a command for causing the text generation model to output a correction proposal for the input document information. In this embodiment, since a conversational text generation model that receives a text as input is used, a sentence or a text in natural language is generated and transmitted as the correction command, which commands the correction of the input document information using the input document information and related document information. The correction unit 14 then receives the sentence or text output by the text generation model as an answer, and obtains the correction proposal.
[0083] The output unit 15 transmits the proposed revision obtained by the revision unit 14 to the user terminal 2. The output unit 15 of this embodiment further transmits the evaluation information obtained by the evaluation unit 12 and the related document information identified by the identification unit 13.
[0084] The database DB registers a plurality of documents and their attributes in association with each other. In this embodiment, the documents stored in the database DB are legal documents created in the past during the process of an agreement, and each document is assigned a status tag indicating the stage of the arrangement adjustment that it specifies. The stages in the arrangement adjustment include, for example, "received" indicating a proposal received from the other party of the contract, "presented" indicating a proposal presented to the other party, "after internal review" indicating a proposal after internal review, and "concluded" indicating the content finally agreed upon with the other party. Since revisions are often exchanged several times between the other party and the contract before the contract is concluded, the number of times of presentation or receipt may also be managed by the status tag.
[0085] It should be noted that the "document" in the present invention is not limited to a legal document that was actually created in the course of an agreement made in the past. For example, a company's own legal document template may be registered as a document in the database DB.
[0086] 4 is an example of information related to a document stored in the database DB in this embodiment. Each item included in the document background information and the contract background information is used as a document attribute in the present invention. The database DB stores document background information linked to the data body indicating the contents of the document. The document background information includes a document ID, a status tag indicating the stage in the adjustment of the contract (agreement) specified by each document, the creation date and time of the document, and information indicating the person who input the document into the database DB and the legal officer in charge of the document, and is linked to the contract background information by a contract ID that specifies the contract background information regarding the contract (agreement) specified by each document. The legal officer information may be entered by the user's selection when the document is registered, or may be specified from the revision history included in the information regarding the revision of the document, the ID of the creator of the comment, etc.
[0087] The contract background information includes information on the contract type, the contracting party (the other party to the contract), the company's position, and the subject of the contract. The contract type is an example of the "type of arrangement" in this invention. Other types of arrangements include atypical contracts, types of rules or regulations, etc. The company's position refers to the contractual position, such as information on the discloser or recipient of information in a confidentiality agreement. The subject of the contract is an example of the "subject of the arrangement," and refers to products, services, or businesses that are the subject of each contract (arrangement). For example, the subject product in a sales contract corresponds to the subject of the contract.
[0088] As described above, multiple documents are often created through repeated revisions before a single contract is concluded. In this embodiment, the document background information for multiple documents created in the process of concluding the same contract is linked to the same contract background information in the database DB. This makes it possible to identify documents relating to the same agreement in this embodiment.
[0089] The data structure can be changed as desired. For example, in addition to the document background information and the contract background information, contract family information for managing a group of mutually related contracts may be defined. Specifically, the document background information may include a contract family ID that identifies the contract family information, and the document background information for multiple documents created in the process of concluding the same contract may be managed by the contract family ID. In this case, the contract family information may include a contract ID that identifies the contract background information, and the document background information may not include a contract ID.
[0090] The user terminal 2 also includes a display unit 21. The user terminal 2 receives the information transmitted by the output unit 15, and the display unit 21 performs display based on the information. Display examples will be described later.
[0091] In addition, each of the above-mentioned units and a part of the database DB of the document editing support device 1 may be provided in a computer other than the document editing support device 1, for example, a user terminal 2 or a generating device 3. In addition, the document editing support device 1 may be provided with a display unit and a generating device.
[0092] <5. Processing Procedure> Next, the process of generating a proposed revision in this embodiment will be specifically described with reference to Fig. 5. Note that the procedure described below is an example, and it goes without saying that it can be modified as appropriate. In the following specification and drawings, it is assumed that the input document and related documents are contracts, and the input document may be referred to as an "input contract" and the related document may be referred to as a "related contract."
[0093] (1) Analysis based on input documents First, in step S501, the user terminal 2 transmits an input document to the document correction support device 1, and the information is acquired by the acquisition unit 11. Note that, although it is assumed here that the entire input document is transmitted, in cases where the user wishes to make corrections only in part, only part of the input document may be transmitted.
[0094] In step S501, in addition to the document file of the input document, attributes of the input document are also transmitted. Here, the attributes of the input document are information about the input document or information about the background of the contract (agreement) stipulated by the input document. For example, the attributes include the other party of the agreement stipulated by the input document, the relationship with the other party such as the contractual position, the type of agreement, the subject of the agreement (contract subject), the creation date and time, the person who input the input document, the legal officer, the requests of the parties regarding the agreement, and the like. The same applies to the attributes of the document registered in the database DB. Furthermore, a status tag indicating the stage in the adjustment of the contract (agreement) stipulated by the input document may be further acquired as an attribute of the input document.
[0095] The attribute may also be term sheet information indicating the contents of a term sheet used in contract negotiations. A term sheet is an arrangement in tabular form of the contents agreed upon with the other party regarding the main items of a contract or the contents of a proposal, and the term sheet information is information including the contents described in the term sheet. The attribute may also be question information based on the question contents and question results generated by question software that asks questions about the contents to be included in a legal document (e.g., a contract).
[0096] In step S502, the identification unit 13 in the document editing support device 1 identifies related documents from among the documents registered in the database DB and generates related document information. Then, in step S503, the priority order of the identified related documents is determined. The details of the process of identifying related documents and determining the priority order will be described later.
[0097] In step S504, the evaluation unit 12 acquires evaluation information. In this embodiment, the evaluation unit 12 executes evaluation of the input document based on the information acquired by the acquisition unit 11. The evaluation of the input document may be performed by any method, such as evaluation by a machine learning model or evaluation by a rule base. Note that the evaluation process in step S504 is not essential and may be omitted.
[0098] When the evaluation unit 12 finishes the evaluation process of the input document, in step S505, the output unit 15 transmits the evaluation information and related document information to the user terminal 2. The related document information is generated for each correction unit corresponding to the input document information based on the related documents, and transmitted. When the user terminal 2 receives the information, the evaluation information and related document information are subjected to display processing, and the processing results are displayed on the display.
[0099] (2) Display of related documents Fig. 6 is an example of a display screen of related documents corresponding to an input document. The related document display screen W6 displays an outline (attributes) of the input document and an outline (attributes) of the related documents at the top of the screen. In the example of Fig. 6, the identification unit 13 has identified two documents as related documents, and an outline of each of them is displayed.
[0100] Here, by selecting the "Change" button, the related documents can be changed to other documents registered in the database DB. When changing related documents, documents registered in the database DB can be searched for based on document background information and contract background information. For example, documents can be narrowed down by status tag, creation date and time, inputter, legal officer, contract type, contract holder, company position, contract target, etc., and the selection of documents to be made related from among them can be accepted. Note that when identifying related documents for each revision unit, i.e., for each input document information, such related documents are displayed for each revision unit, and related documents can also be changed for each revision unit.
[0101] The contents of the input document specified by the information sent in step S501 are displayed on the left side of the screen. Also, a "Edit read contents" button is provided, and if there is an error in the contents specified from the information sent in step S501, the contents can be edited by selecting this button.
[0102] On the right side of the screen, related document information is displayed for each input document information (in this embodiment, for each clause or group of clauses of the input document). Here, for each related document information, the contract name included in the contract background information (such as "Company A NDA1" in FIG. 6), the contents of the related document, information regarding the revisions made when creating the related document, the creation date or upload date of the related document, the status tag ("Presented ver." or "Received ver." in FIG. 6), and a revision button are displayed.
[0103] The related document display screen W6 includes a highlight selection section W61, which allows the user to set whether to highlight the differences or matching parts of the related document information with the input document (input document information). Note that highlighting may be performed on the differences or matching parts between the contents before and after the correction in the correction history of the related document information. Here, the related document information in this embodiment includes information regarding corrections made when the related document was created.
[0104] The information on the revision is, for example, information indicating the content of the revision made when the related document was created. The information on the revision may also include other information indicating the purpose of the revision, the time of the revision, the person who made the revision, etc. The information on the revision may be, for example, information indicating the revision location in the document registered in the database DB, or information based on the change history function provided by the document creation software. The information on the revision preferably includes the content before and after the revision in the revision location, and may further include information indicating whether the revision was a deletion, addition, or movement, the date and time of the revision, and information on the user who made the revision.
[0105] In this embodiment, as described above, multiple documents created in the process of one contract are linked to each other and registered in the database DB. Based on this information, and based on the differences between multiple documents related to the same contract, the identification unit 13 or the correction unit 14 may generate information regarding the correction. Then, in step S505, the output unit 15 transmits information regarding the correction of the document identified as the related document to the user terminal 2, and displays it on a screen such as that shown in FIG.
[0106] 6, the contents of the document before and after the revision are displayed as information related to the revision, and the parts that have been changed before and after the revision are displayed in a different form from the parts that have not been changed. In this embodiment, related document information is displayed for each clause or group of clauses.
[0107] Specifically, the parts related to the changes are highlighted. Here, of the documents related to contracts that have many points in common with the attributes of the input document with regard to the contract background information registered in the database DB, documents that have been given a presented tag as a status tag are identified as related documents. Then, as the document immediately preceding that related document, the contents of a document that has been given a received tag among documents related to the same contract are displayed as part of the information regarding the revision. Note that both of these may be identified as related documents.
[0108] A correction button W62 is arranged on the display portion of each related document information. When the user operates the correction button W62, the user terminal 2 transmits a correction request in step S506 to the document correction support device 1. The correction unit 14 may also acquire a correction proposal based on each of a plurality of related document information, and when the user selects "correct with all examples" at the top of each clause, a correction request requesting acquisition of a correction proposal based on each related document information is transmitted in step S506. In addition, a "create a correction proposal with recommended related contract" button is displayed for each correction unit, and when the user selects this button, related document information of the related document with the highest priority described later for the correction unit is selected, and a correction request requesting acquisition of a correction proposal based on each related document information for each correction unit is transmitted in step S506. Note that a button for performing correction in bulk may be further provided, and a correction request requesting acquisition of a correction proposal for multiple correction units in bulk may be transmitted.
[0109] In this embodiment, multiple related documents are identified, and related document information for each of them is displayed. The display order here is determined based on the priority order described below. Specifically, the higher the priority order, the higher the document is displayed. Also, only the related document information with the highest priority order may be displayed, or a pre-set number of related document information may be displayed in descending order from the top. There is no restriction on the display format, and multiple related document information may be displayed horizontally instead of vertically. Also, the full text of multiple related documents may be displayed horizontally or vertically.
[0110] In addition, if the evaluation unit 12 acquires evaluation information in step S504, the evaluation information can also be displayed. Fig. 7 shows an example of the display of an evaluation information display screen W7 that displays evaluation information together with related document information. The main difference from Fig. 6 is the configuration on the right side of the screen.
[0111] In this case, the evaluation results based on the evaluation information are displayed on the right side of the screen. For each issue, the issue contents W71, including a type and explanation indicating whether the issue is a clause deletion, addition, or content modification, and the related document display section W72 are displayed. In this case, the correction unit is each issue. The display content in the related document display section W72, which displays the related document information, is the same as in FIG. 6. Note that FIG. 7 shows an example in which the corrected parts are simply highlighted as information regarding the corrections made when creating the related documents.
[0112] (3) Generation of revision proposals When the revision request is sent in step S506 by the operation via the screen as described above, the revision unit 14 receives it and executes a process for acquiring a revision proposal. Specifically, the acquisition unit 11 acquires the input document information received in step S501 and the related document information specified in the revision request.
[0113] The correction unit 14 inputs the input document information and related document information including information on corrections made when the related document was created to the generation device 3, and causes the generation device 3 to generate a correction proposal. Specifically, the correction unit 14 generates a correction command including the input document information, related document information indicating the contents of the related document, and information on corrections made when the related document was created.
[0114] In this embodiment, the information on the correction includes the description before and after the correction, the correction parts, the user who corrected each correction part, the comments on the correction, and the user who wrote the comments. The correction unit 14 generates a correction command based on the information on the corrections and comments made by a user in the same position as the user who sent the input document information, among the information on the corrections.
[0115] The comments on the revisions are preferably associated with parts of the document, and may include comments for the user's own company (or for the user's own client, if the user is a legal adviser for a contracting party) and comments for the contracting party.
[0116] Note that "the same position as the user who sent the input document" and "the same position as the user who is the inputter" refer to the users on the company's side. That is, among the parties to the contract, it refers to the position of the party on the same side as the inputting user or the position of supporting that party. Specifically, users in the same position include employees and officers of the same company, lawyers or law firms on the same side, etc.
[0117] Here, in this embodiment, in generating the correction instruction, in addition to the input document information and the related document information, the attributes of the input document, evaluation information, requests regarding correction, etc. may be further considered. In this case, the acquisition unit 11 further acquires the attributes of the input document. In this embodiment, the input document received in step S501 and its attributes are stored in the storage unit of the document correction support device 1, and the acquisition unit 11 acquires the input document information and its attributes based on them. Regarding the evaluation information, the evaluation information acquired in step S504, and regarding the requests regarding correction, the information received as a correction request in step S506 are stored in the storage unit of the document correction support device 1, and the acquisition unit 11 acquires them respectively at the time of creating the correction instruction in step S507 and passes them to the correction unit 14.
[0118] The correction unit 14 generates a correction instruction for each correction unit. In this embodiment, when a correction instruction is input via a screen that displays related document information for each clause or group of clauses as shown in FIG. 6, the displayed clause or group of clauses is taken as the correction unit. Also, when a correction instruction is input via a screen that displays evaluation information as shown in FIG. 7, a part of the input document associated with the pointed-out matter is taken as the correction unit.
[0119] Here, for example, if the correction unit is a group of clauses, if there is a clause included in multiple groups, the clause may be corrected twice, and an appropriate correction proposal may not be created. Therefore, in such a case, a correction command may be generated so that clauses included in multiple groups are not subject to correction. In addition, the correction unit 14 may generate a correction command to instruct the correction of clauses included in other groups to generate a correction proposal taking into account the corrections in the other groups. For example, the correction unit 14 may generate a correction command that includes input document information corresponding to one correction unit and multiple related document information corresponding to the one correction unit, and generates a correction proposal for the input document information based on the multiple related document information.
[0120] The sentence generation model of this embodiment is a conversational model that receives a sentence in a natural language as input and generates a sentence as a response, so the correction command is generated as a sentence in a natural language. The generation of the modification command will now be described in detail.
[0121] First, for attributes of the input document, words expressing each attribute can be added as is. Also, for requests regarding correction, when the requests are input by selection, first, the words corresponding to the selected request are identified. For example, by setting corresponding words in advance for each option of requests, the words corresponding to the selected request can be identified. Then, the correction unit 14 generates a correction command based on those words, the input document information, the related document information, and a command statement separately registered in advance in the storage unit of the document correction support device 1.
[0122] Here, it is preferable to store a character string describing the rules for document creation as the directive. For example, in legal documents, the legal meaning may differ between describing an exception in a sentence of a provision and describing a disclaimer separately after the sentence. By combining words that specify the rules for such description in response to requests for revisions, etc., appropriate revision proposals can be generated.
[0123] As the instruction sentence, for example, "Please revise the input contract sentence with reference to the following sentence before and after revision as revision examples," "The following comments were made about the above clause as a result of the review. Please create a revision proposal based on the comments and with reference to the information on the descriptions and revisions of the related documents below," or "Please create an appropriate contract sentence based on the following rules with reference to the information on the descriptions and revisions of the related documents for the above clause," etc. can be stored in advance in the storage unit of the document revision support device 1. Then, a sentence that combines the input document information, the related document information, the instruction sentence, and words based on the attributes and requests may be generated as a revision command for instructing the sentence generation model to generate a revision proposal.
[0124] The correction unit 14 may generate a plurality of correction instructions so as to generate a correction proposal using each of a plurality of related document information for one input document information. Also, the correction unit 14 may generate a correction instruction instructing to propose a plurality of correction proposals using one related document information for one input document information.
[0125] Furthermore, the modification unit 14 may generate a modification command to cause the generating device 3 to generate comments indicating viewpoints and reasons for the modification, and the comments may be acquired in addition to the proposed modification. Furthermore, the comments may be generated and acquired as comments for the user's own company (or as a reason for the modification for the user's client in the case where the user is a legal advisor for a contracting party, etc.) and comments for the contracting party.
[0126] In contract negotiations, both parties make repeated revisions to the contract, and at this time, it is common to write at least one of comments for internal explanation and comments for external explanation. Writing comments requires a high level of expertise, but the above configuration can help even users without specialized knowledge to write comments.
[0127] Once the revision command has been generated, the process proceeds to step S508, where the revision unit 14 transmits the generated revision command to the generation device 3. The transmission of the revision command to the generation device 3 in step S508 corresponds to the input of the input document information and related document information to the generation device 3. The revision commands for each revision unit generated as described above may each be transmitted in multiple batches. For example, the system may be configured to first transmit the input document information and attributes of the revision command, then transmit the instruction text and related document information, and then transmit the revision information and words according to the request. Alternatively, the system may be configured to determine whether to transmit the revision command all at once or in multiple batches depending on the number of characters in the revision command.
[0128] When the generation device 3 receives the correction command, it inputs the correction command to the sentence generation model. As a result, the sentence generation model outputs a correction proposal corresponding to the input document information (step S509). In this embodiment, the correction unit 14 generates a correction command for each correction unit, and in step S509, a correction proposal is generated for each correction unit. Also, multiple correction proposals may be generated for each correction unit in response to the correction command.
[0129] In step S510, the generating device 3 transmits the proposed revision to the document revision support device 1. In the document revision support device 1, the revision unit 14 acquires the proposed revision, and in step S511 transmits it to the user terminal 2. The user terminal 2 displays the proposed revision based on the received information.
[0130] (4) Display of proposed amendments Next, a display example of the proposed revision on the user terminal 2 will be described with reference to Fig. 8 and Fig. 9. Fig. 8 is a display example in which only the proposed revision is displayed, and Fig. 9 is a display example in which the proposed revision is displayed together with the evaluation information. Here too, the proposed revision is displayed for each revision unit, similar to the related document display screen W6 and the evaluation information display screen W7.
[0131] The proposed revision display screen W8 in Fig. 8 displays proposed revisions W81 on the right side of the screen. When proposed revisions are generated based on multiple pieces of related document information, multiple proposed revisions are displayed for each revision unit. Here, an example is shown in which the revision unit is a clause or a group of clauses. In addition, a copy button is provided for each proposed revision, and the content of the proposed revision can be copied to the clipboard of the user terminal 2 by the user operating the copy button.
[0132] When the document correction assistance program of the present invention is provided as an add-on (add-in, plug-in) program for document editing software, a correction button may be provided instead of the copy button. In this case, when the user operates the correction button, corrections are made directly on the document editing software using the correction suggestions.
[0133] The revision proposal display screen W9 in Fig. 9 displays a revision proposal W91 on the right side of the screen. Here too, the revision proposal W91 is displayed for each revision unit. In this case, the revision unit is a part of the input document that corresponds to the pointed out item, and evaluation information is displayed together with the revision proposal W91.
[0134] As described above, according to the present embodiment, it is possible to provide an appropriate revision proposal for an input document by taking into consideration the contents of the input document and related documents. In addition, by generating revision proposals using a sentence generation model, it is possible to provide natural revision proposals that reflect the context and conditions of the input document.
[0135] <6. Related Document Identification Processing> The present invention also relates to a technology for appropriately identifying related documents related to an input document. The technology described in Patent Document 1 exemplified in the Background Art merely identifies contracts that are similar in content as similar contracts, and therefore has a problem in that it is unable to effectively support revisions based on exchanges of revisions during the contract process. In another aspect of the present invention, this is a problem to be solved.
[0136] The process of identifying related documents and determining priority order in steps S502 and S503 in the above-described embodiment will be described in detail below. Note that the following description is an example and may be modified as desired. For example, a legal viewpoint (such as an exception to confidential information) included in the input document information may be identified using a rule base or machine learning model, and documents including clauses of the same legal viewpoint may be identified as related documents. In addition, the present invention may not require the generation of the above-described proposed revision, and may assist the user in revising the document by identifying related documents corresponding to the input document and providing them to the user.
[0137] 10 is a flowchart showing a procedure of a process in which the identification unit 13 determines related documents and determines priorities. First, in step S101, the identification unit 13 acquires the attributes of the input document acquired by the acquisition unit 11 together with the input document information in step S501.
[0138] Then, in steps S102 to S104, the identification unit 13 identifies related documents. In this embodiment, first, in step S102, the identification unit 13 searches for contract background information registered in the database DB that matches the attributes based on the attributes related to the contract specified by the input document. The attributes of the input document used here include information similar to the items of the contract background information, such as the other party to the agreement, the relationship with the other party such as contractual position, the type of agreement, and the requests of the parties regarding the agreement, in the contract specified by the input document.
[0139] Next, in step S103, the identification unit 13 identifies related contracts according to the match between the attributes of the input document and the contract background information. Specifically, among multiple attributes, those with the greatest number of matching attributes are identified as related contracts. For example, a predetermined number of matching attributes may be identified as related contracts in descending order of the number of matching attributes, or all contracts in the contract background information with a predetermined number or more matching attributes may be regarded as related contracts.
[0140] In step S104, the identification unit 13 identifies a related document in accordance with a match between the document background information linked to the contract background information of the related contract identified in step S103 and the attributes of the input document. The attributes of the input document used here include information similar to the items of the document background information, such as the creation date and time, the person who input the input document, the legal officer, etc.
[0141] Specifically, the identification unit 13 identifies, based on the attributes of the input document, documents corresponding to the document background information linked to the contract background information of the related contract that have the largest number of matching attributes as related documents. For example, a predetermined number of documents may be identified as related documents in descending order of the number of matching attributes, or all documents related to document background information that have a predetermined number or more matching attributes may be regarded as related documents. At this time, the status tag may also be taken into consideration, but in this embodiment, the status tag is not taken into consideration when checking the matching of attributes.
[0142] In the case where the upper limit number of related documents to be identified is determined in advance, when there are multiple pieces of document background information with the same number of matching attributes, the identification unit 13 identifies related documents based on status tags. Specifically, documents to which a presented tag indicating that a party in the same position as the user who inputted the input document has presented the document to the other party of the contract (documents corresponding to document background information with the status tag set to "presented") are given priority and identified as related documents.
[0143] Also, the related document to be prioritized may be determined according to the status tag of the input document. For example, a document to which a status tag indicating the next stage of the stage indicated by the status of the input document is attached may be preferentially identified as a related document. Specifically, if the status tag of the input document indicates that it is at the "first receipt" stage, a document to which a status tag indicating the next stage, "first presentation" (or "second presentation" if the first party has presented it first), may be preferentially identified as a related document.
[0144] Alternatively, a document to which a status tag indicating the same stage as the status tag of the input document is attached may be preferentially identified as a related document. Specifically, if the status tag of the input document indicates that the document is at the "first receipt" stage, a document to which a status tag indicating the same stage as the "first receipt" may be preferentially identified as a related document.
[0145] When a document with a status tag indicating the next stage of the input document is given priority, the user can consider amending the input document based on the amendments made by the user's company in the previous agreement using the related documents. Also, when a document with a status tag indicating the same stage as the input document is given priority, the user can consider amendments by checking the amendments presented by the other party in the previous agreement using the related documents.
[0146] Note that "in the same position as the user who sent the input document" or "in the same position as the user who inputs the document" refers to a user on the same company's side. In other words, it refers to a party to the contract who is on the same side as the user who inputs the document, or a party who supports the party. Specifically, a user in the same position includes employees or executives of the same company, allied lawyers or law firms, etc.
[0147] In addition, when there are multiple documents with the same number of matching attributes and tagged with the "present" tag, the identifying unit 13 of this embodiment identifies related documents based on the time when each document was created. Specifically, when there are multiple documents with the present tag, the identifying unit 13 preferentially identifies, as related documents, documents created at an earlier stage in the agreement stipulated by each document.
[0148] For example, if document A is the first document presented to the other party in a contract, and document B is the second document presented to the other party in a contract, both may be tagged as "presented." In such a case, the identifying unit 13 prioritizes document A, which is presented at an earlier stage, i.e., the first time it is presented, and identifies it as a related document.
[0149] Here, the present invention may be in a form in which the status tag of the document is not stored in the database DB. In this case, among documents relating to the same agreement, the contract that is the second oldest chronologically (registration date and time, creation date and time, etc.) can be regarded as the related contract. This is because the document received from the other party is often the oldest, and the second oldest document is often the document that one company presented to the other party.
[0150] In this way, in a contract, the document first presented to the other party often most strongly reflects the intentions of the company, so documents created at an earlier stage in each agreement, especially documents first presented to the other party by the company, may be given priority as related documents. Note that it is preferable to configure the system so that the user can select and set which status tag is given to which document to give priority.
[0151] Here, the identification unit 13 may identify related documents for each correction unit, i.e., for each input document information. As described above, in this embodiment, since input document information is generated for each correction unit, multiple pieces of input document information may be obtained for one input document. In such a case, the identification unit 13 may execute step S104 for each piece of input document information and identify different related documents for each piece of input document information. Specifically, for example, if the revision unit is each clause, the relevant document related to the input document information of a certain clause may be document A, while the relevant document related to the input document information of another clause may be document B.
[0152] In this embodiment, the specification unit 13 specifies a plurality of related documents in the above-mentioned procedure, but the procedure for specifying related documents may be changed arbitrarily. For example, only one related document may be specified. In the present embodiment, in steps S102 to S104, the procedure is described in which the related contract is first identified and then the related document is identified, but the present invention is not limited to this. For example, the identification unit 13 may check for matches with the attributes of the input document without particularly distinguishing between the contract background information and the document background information for each document, and identify documents with a large total number of matching items as related documents.
[0153] Next, the determination of the priority order of the multiple related documents in step S105 will be described. The identification unit 13 determines the priority order for the multiple related documents. The priority order is taken into consideration when displaying related document information on the related document display screen W6 and the evaluation information display screen W7 described above, and when generating a revision proposal when the user does not select a related document to be used in generating a revision proposal.
[0154] The identification unit 13 determines the priority order for multiple related documents based on the matching of attributes with the input document, status tags, and the time of creation. For attributes, the priority order is determined so that a document having a greater number of matching attributes with the input document in the document background information and contract background information is given a higher priority. Also, a weight or rank may be set for each attribute so that a document having a matching attribute of greater importance is given a higher priority.
[0155] Regarding the status tags, the specification unit 13 determines the priority order so that the document to which the presentation tag is added (the document corresponding to the document background information in which "presentation" is set as the status tag) has a higher priority order.
[0156] Also, the related document to be prioritized may be determined according to the status tag of the input document. For example, the priority order may be determined so that the priority order of the related document to which a status tag indicating the next stage after the stage indicated by the status of the input document is assigned is higher. Specifically, if the status tag of the input document indicates that it is at the "first receipt" stage, the priority order may be determined so that the priority order of the related document to which a status tag indicating the next stage, "first presentation" (or "second presentation" if the first party has submitted the document first) is assigned is higher.
[0157] Alternatively, the priority order may be determined so that a related document to which a status tag indicating the same stage as the status tag of the input document is attached has a higher priority. Specifically, if the status tag of the input document indicates that the document is at the "first receipt" stage, the priority order may be determined so that a related document to which a status tag indicating the same stage as the "first receipt" stage is attached has a higher priority.
[0158] When a related document with a status tag indicating the next stage of the input document is given priority, the user can consider amending the input document based on the amendments made by the user's company in the previous agreement using the related document. Also, when a related document with a status tag indicating the same stage as the input document is given priority, the user can consider amendments by checking the amendments presented by the other party in the previous agreement using the related document.
[0159] Regarding the creation time, the specifying unit 13 determines the priority order so that a document created at an earlier stage in the contract that each document specifies has a higher priority order. When there are multiple related documents to which the "presented" tag is attached, the specifying unit 13 may determine the priority order so that a document created at an earlier stage in each agreement has a higher priority order. In particular, when there is a related document that was first presented to the contracting party by a party in the same position as the user who transmitted the input document, the specifying unit 13 may determine the priority order so that the related document has the highest priority order. Note that it is preferable to configure the system so that the user can select and set which status tag is attached to which the related document has a higher priority order.
[0160] In addition, if there are multiple related documents that still have the same priority even after determining the priority using the above procedure, the identification unit 13 may determine the priority so that, for example, the related document with the most recent creation date is given the higher priority.
[0161] By determining the priority order in this way and reflecting it in the display of related document information, the user can efficiently obtain related document information that is more useful for correcting the input document. In addition, in determining the priority order, by giving priority to the document that the company first presented, particularly in past contracts, the related document that better reflects the company's intentions is given priority, improving convenience.
[0162] As described above, according to the present invention, it is possible to provide an appropriate correction proposal for an input document based on descriptions in the input document and related documents. Furthermore, according to another aspect of the present invention, it is possible to effectively identify related documents that are more easily used as references for correcting an input document. Furthermore, by presenting the identified related documents to a user, it is possible to effectively support the work of correcting an input document. [Explanation of symbols]
[0163] 0: Document revision support system 1: Document editing support device 11: Acquisition part 12: Evaluation section 13: Specific part 14: Correction section 15: Output section 2: User terminal 21:Display section 3:Generation device 10: Information processing device 101: Control unit 102: Storage section 103: Communications Department 90: Terminal equipment 901: Control unit 902: Storage section 903: Communications Department 904: Input section 905: Output section NW: Network W6: Related documents display screen W61: Highlight selection section W62:Edit button W7: Evaluation information display screen W71: Content of the comment W72: Related Documents Display Section W8: Amendment display screen W81: Amendment proposal W9: Amendment display screen W91: Amendment proposal
Claims
1. A document correction assistance program for assisting in document correction, comprising: A computer is caused to function as an acquisition unit, a correction unit, and an output unit; The acquisition unit acquires input document information indicating at least a part of the content of an input document, which is a legal document to be amended, and related document information indicating at least a part of the content of a related document, which is a legal document related to the input document; the correction unit inputs the input document information and the related document information to a generation device, causes the generation device to generate a proposed correction for the input document information, and obtains the proposed correction; The output unit transmits the proposed revision.
2. The related document information further includes information regarding modifications made in creating the related document; The document correction assistance program according to claim 1 , wherein the correction section further inputs information about the correction of the related document to the generating device to generate the proposed correction.
3. 3. The document correction support program according to claim 2, wherein the information regarding the correction includes a comment indicating the content of the correction in the related document.
4. 3. The document correction support program according to claim 2, wherein the information regarding the correction includes information indicating contents of at least a part of the related document before and after the correction.
5. 5. The document correction assistance program according to claim 4, wherein the information regarding the correction further includes information regarding a correction portion identified based on a difference between contents of the related document before and after the correction.
6. 3. The document correction support program according to claim 2, wherein the information regarding the correction includes information regarding a correction portion in the related document.
7. The information regarding the correction includes, for each of the corrections, information about the user who made the correction; 7. The document correction support program according to claim 5, wherein the correction unit causes the generation device to create the correction proposal based on information about corrections made to the related document by a user in the same position as the user who transmitted the input document information.
8. The generation device is a device for generating the revision suggestion using a sentence generation model, The document correction assistance program according to claim 1 , wherein the correction unit generates a correction command in natural language based on the input document information and the related document information, and transmits the command to the generation device, thereby generating the proposed correction.
9. the input document information is information indicating the contents of one of the correction units when the input document is divided into predetermined correction units, The acquisition unit acquires a plurality of pieces of input document information for one of the input documents, The document correction assistance program according to claim 1 , wherein the correction section causes the generating device to generate the proposed correction for each of the input document information.
10. the input document and the related document are legal documents containing a plurality of clauses; 10. The document correction support program according to claim 9, wherein the correction unit is a clause or a group of a plurality of clauses defining a set of rules contained in the input document or the related document.
11. The computer further functions as an evaluation unit; The evaluation unit acquires evaluation information including a pointed out item associated with a portion of the input document, 10. The document correction support program according to claim 9, wherein the correction unit is a part of the input document corresponding to the indicated matter.
12. The computer further functions as a specific unit, The acquisition unit further acquires attributes of the input document, 2. The document correction support program according to claim 1, wherein the identifying unit identifies, as the related document, a document having commonality with the input document based on attributes of the input document from a database in which a plurality of documents and their attributes are registered.
13. The input document is a legal document relating to an agreement with a specific party; 13. The document correction support program according to claim 12, wherein the attributes of the input document and the attributes of the document registered in the database include any one of the other party of the agreement, the user's position in the agreement, the type of the agreement, and the target of the agreement.
14. The identification unit identifies a plurality of related documents, The document correction assistance program according to claim 12 , wherein the correction unit transmits the related document information based on a user's selection from the identified plurality of related documents to the generation device to generate the proposed correction.
15. The database stores, as documents, legal documents that have been created in the past during the course of the arrangement, and stores, for each of the documents, a status tag that indicates a stage in the adjustment of the arrangement that each document specifies; The document correction assistance program according to claim 12 , wherein the identifying section identifies the related document based on the status tag.
16. The identification unit identifies a plurality of documents related to the same agreement to which different status tags are assigned as the related documents, 16. The document correction assistance program according to claim 15, wherein the correction unit transmits information about corrections to the related documents, generated based on differences between the plurality of documents, to the generation device to generate the proposed correction.
17. 16. The document correction support program according to claim 15, wherein the identification unit preferentially identifies as the related document a document to which a presentation tag has been assigned as the status tag indicating that the document has been presented to the other party in the agreement by a party in the same position as the user who sent the input document in the adjustment of the agreement.
18. 20. The document correction support program according to claim 17, wherein when there are a plurality of documents to which the presentation tag is assigned, the identification unit prioritizes identifying, as the related document, the document that was created at an earlier stage in an agreement specified by each document.
19. 17. The document correction assistance program according to claim 16, wherein the identifying unit identifies a plurality of the related documents and determines priorities of the plurality of the related documents such that a document to which the presentation tag is added has a higher priority.
20. 20. The document correction support program according to claim 19, wherein, when there are multiple documents to which the presentation tag is assigned, the identification unit determines the priorities of the multiple related documents such that a document created at an earlier stage in an agreement specified by each document has a higher priority.
21. 20. The document correction assistance program according to claim 19, wherein the correction section inputs the related document information relating to a related document having the highest priority among the plurality of related documents to the generation device to generate the proposed correction.
22. The computer further functions as a display unit; 20. The document correction support program according to claim 19, wherein said display unit displays said related document information relating to a plurality of said related documents in order of priority.
23. The database stores a plurality of legal documents created in the past in the course of the same agreement together with information indicating a time series of the documents; 18. The document correction support program according to claim 17, further comprising: a step of giving priority to a document that is older in chronological order from among the documents related to the same agreement to which the presented tag is added, and identifying the document as the related document.
24. The correction unit further obtains a comment indicating a correction content for the input document, The document correction assistance program according to claim 1 , wherein the output unit further transmits the comment.
25. A document correction support system for supporting document correction, comprising: An acquisition unit, a correction unit, and an output unit, The acquisition unit acquires input document information indicating at least a part of the content of an input document, which is a legal document to be amended, and related document information indicating at least a part of the content of a related document, which is a legal document related to the input document; the correction unit inputs the input document information and the related document information to a generation device, causes the generation device to generate a correction proposal for the input document information, and obtains the correction proposal; The output unit transmits the proposed revision.
26. A document correction assistance method for assisting in correcting a document, comprising: Obtaining input document information indicating at least a portion of the content of an input document, which is a legal document to be amended, and related document information indicating at least a portion of the content of a related document, which is a legal document related to the input document; inputting the input document information and related document information into a generating device, causing the generating device to generate a revision proposal for the input document information, and obtaining the revision proposal; The document correction assistance method further comprises transmitting the proposed correction.
Citation Information
Patent Citations
Document processing program and information processing device
WO2020230346A1