Information processing method, information processing device, and information processing system

Abstract

Provided is an assistance technique for achieving more efficient review of a document. An information processing method according to one embodiment comprises a process in which a computer identifies a corrected portion in a document with which assistance is to be given and outputs at least one piece of alert information for the corrected portion in the document with which assistance is to be given, on the basis of a first standard having one or a plurality of pieces of alert information for drawing the attention of a user.

Description

Information Processing Method, Information Processing Apparatus, and Information Processing System

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

[0002] Conventionally, systems for supporting document review have been proposed. For example, there has been proposed a legal document evaluation method implemented by a computer for evaluating a legal document, the method including: an input step of receiving an input of the legal document; a reading step of reading out evaluation range information relating to an evaluation range of the legal document and an evaluation rule for the legal document; an evaluation step of evaluating at least a range corresponding to the evaluation range information in the legal document based on the evaluation rule; and an output step of outputting the evaluation (Patent Document 1).

[0003] Japanese Unexamined Patent Application Publication No. 2019-212315

[0004] A problem to be solved by the present disclosure is to provide a support technique that realizes more efficient document review.

[0005] In the information processing method according to the embodiment, a computer identifies a corrected portion in a support target document, and outputs at least one piece of alert information to the corrected portion of the support target document based on a first criterion having one or more pieces of alert information that call a user's attention.

[0006] Figure 1 is a schematic diagram illustrating the outline of a document viewing support system according to an embodiment. Figure 2 is a block diagram showing an example of the hardware configuration of an information processing device according to an embodiment. Figure 3 is a diagram showing an example of functional blocks in the processor and storage device of the information processing device according to this embodiment. Figure 4A is a diagram showing an example of alert display of review information provided to the user in the terminal device of this embodiment. Figure 4B is a diagram showing an example of alert display of review information provided to the user in the terminal device of this embodiment. Figure 5 is a block diagram showing an example of the hardware configuration of the terminal device of this embodiment. Figure 6 is a diagram showing an example of functional blocks in the processor and storage device of the terminal device of this embodiment. Figure 7 is a flowchart showing the flow of document viewing support processing including review processing according to the first example of this embodiment. Figure 8 is a flowchart showing the flow of review processing according to the first example of this embodiment.

[0007] The embodiments will be described in detail below with reference to the drawings. In the following description, components having substantially the same function and configuration will be denoted by the same reference numerals, and will be described only when necessary. Furthermore, the embodiments can be combined with other embodiments, modifications, and prior art.

[0008] <Schematic Diagram> Figure 1 is a schematic diagram illustrating the outline of the document viewing support system S (an example of an information processing system) according to this embodiment. As shown in Figure 1, the document viewing support system S of this embodiment functions as a support system when a user views a document, and includes, for example, an information processing device 10 according to the embodiment, and at least one terminal device T1 to TN that can communicate with the information processing device 10 via a network N. The terminal devices T1 to TN are operated, for example, by users P1 to PN, respectively. It also includes a server 50 equipped with a Large Language Model (LLM). In Figure 1, the information processing device 10 and the server 50 are implemented as separate devices, but the information processing device 10 and the server 50 may be implemented as a single device (i.e., the information processing device 10 may be equipped with an LLM). The configuration of terminal device T1 will be described as an example of the terminal devices T1 to TN.

[0009] Here, the document viewing support system S according to this embodiment is, for example, a system that supports the viewing (display) of documents on a browser. In this specification, "document" means any document in any language that is created in accordance with certain rules or standards, such as a rule, law, or custom. "Document data" is data of a document, regardless of its data format (document data is also simply referred to as "document"). Document data is typically text data, but can also include data such as image data and tabular data. Furthermore, a document or document data may include metadata (version, creation date, update date, creator, updater, etc.), formatting information (decorations such as underlines and markers, fonts, indents, etc.), change history information, comment information, etc., created by word processing software. Note that "version" may refer to the file version (edition) or the software version.

[0010] In the following explanation, to make the details clearer, we will use the example of a legal document (e.g., a contract) written in Japanese that is handled by the document viewing support system S. However, the documents handled by the document viewing support system S are not limited to contracts; they may also include, for example, internal organizational regulations, or even company regulations in English. Furthermore, the legal documents handled by the document viewing support system S may be written in any language.

[0011] Furthermore, the document viewing support system S according to this embodiment may also include a document creation support function in addition to the document viewing support function. Here, the document creation support function refers to a function for editing (modifying, adding to, deleting, etc.) the content of the document to be supported, and providing information to support editing. In the following, to make the explanation more concrete, we will use the case where the document viewing support system S includes a document creation support function as an example.

[0012] The services provided by the document viewing support system S according to this embodiment (document viewing support services) may include, for example, at least one of the following: performing document review processing on a document to be supported on a web browser and providing the results to the user; performing document editing in response to instructions from the user; saving or providing the edited document to the user; editing additional information associated with the document to be supported (e.g., metadata, comment information, chat information, formatting information, etc.); creating a new document (including a copy); comparing multiple documents (e.g., highlighting difference information); automatically proofreading the document to be supported; and performing various processes using a translation display function.

[0013] Furthermore, in this specification, document review processing (also referred to as review processing) is a process that analyzes or evaluates the information contained in a document to be supported based on certain rules or criteria and outputs the results. For example, document review processing includes checking and examining whether the information contained in a document to be supported is appropriate based on certain rules or criteria. Specifically, this includes, for example, comparing the wording in the document with a pre-prepared checklist (a playbook is an example of a checklist) and outputting the results of the comparison or the points included in the checklist. Editing, on the other hand, means, for example, adding, deleting, or changing the information contained in the document to be edited.

[0014] Metadata associated with a supported document includes, for example, at least one of the following: document ID, document type (e.g., basic transaction agreement, confidentiality agreement, outsourcing agreement, etc.), version, creation date and time, update date and time, upload date and time, creator, updater, deadline, and document ID of related documents.

[0015] The comment information associated with the supported document in this embodiment is additional information created by at least one user involved in the creation of the document, and is associated and managed, for example, for each version of the supported document. The comment information associated with the supported document may also be comments created by word processing software.

[0016] The chat information in this embodiment refers to additional information generated using a chat function among multiple users involved in the creation of a document to be supported, and is managed, for example, in association with multiple versions of the document to be supported.

[0017] These additional information are typically text data, but can also include other data formats such as images and audio.

[0018] The document supported by this embodiment may be a document file uploaded from a terminal device, a file newly created in an information processing device using an online editor, or a file that has already been uploaded, saved, or edited.

[0019] In this embodiment, the terminal device T1 is a client device managed by user P1 who uses the document viewing support service. The terminal device T1 displays a document on the screen of the display device based on information from, for example, the information processing device 10. In this embodiment, the user can use the document review results, editing screen, etc., displayed on the screen of the display device of the terminal device T1 to perform operations such as indentation replacement processing, document review referencing, comment input, comment referencing, chat information input, chat information referencing, document editing processing, translation display of the supported document, display of document review information, display of difference information, display of suggested sentences, display of points raised, and document saving.

[0020] Next, the specific configuration of the information processing device 10 and terminal device T1 of this embodiment will be described. (Information Processing Device 10)

[0021] The information processing device 10 in this embodiment is connected to a client device via communication and performs various processes in response to requests from the client device. For example, it is a server device managed by a provider of document viewing support services. The information processing device 10 responds to a document viewing support request received from a terminal device T1, which acts as a client device, via the network N, and provides the document viewing support service to the terminal device T1. In particular, the information processing device 10 in this embodiment performs actions such as outputting alert information for modified parts of the document to be supported in the document viewing support service.

[0022] <Hardware Configuration (Information Processing Device)> Figure 2 is a block diagram showing an example of the hardware configuration of the information processing device 10 according to this embodiment. As shown in Figure 2, the information processing device 10 of this embodiment is a computer and, as an example, includes a processor 11, main memory 13, auxiliary memory 15, network interface 17, device interface 18, and a bus 19 that connects each component.

[0023] The information processing device 10 of this embodiment shown in Figure 2 has one of each component, but it may have multiple identical components. Also, although Figure 2 shows one information processing device 10, the software may be installed on multiple computers, and each of the multiple computers may execute the same or different parts of the software. In this case, it may be a distributed computing configuration in which each computer communicates via a network interface 17 or the like to execute the processing. In other words, the information processing device 10 in this embodiment may be configured as a system that realizes the various functions described later by having one or more computers execute instructions stored in one or more storage devices.

[0024] The various calculations of the information processing device 10 in this embodiment may be performed in parallel using one or more processors, or using multiple computers connected via a network. Alternatively, the various calculations may be distributed to multiple processing cores within a processor and executed in parallel. Furthermore, some or all of the processing and means of this disclosure may be performed by at least one of a processor and a storage device located on a cloud that can communicate with the information processing device 10 via a network. Thus, the processing of the document viewing support system S in this embodiment may take the form of parallel computing using one or more computers.

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

[0026] The processor 11 performs calculations based on data and software (programs) input from each of the internal components of the information processing device 10, and can output calculation results and control signals to each of the internal components. The processor 11 may also control each component of the information processing device 10 by executing the OS (Operating System) or applications of the information processing device 10.

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

[0028] In this embodiment, the main memory 13 is a storage device that stores instructions executed by the processor 11 and various data, and the information stored in the main memory 13 is read by the processor 11. The auxiliary storage device 15 is a storage device other than the main memory 13. These storage devices refer to any electronic component capable of storing electronic information, and may be semiconductor memory. The semiconductor memory may be either volatile memory or non-volatile memory. The storage device for storing various data used in the information processing device 10 according to this embodiment may be realized by the main memory 13 or the auxiliary storage device 15, or by the built-in memory of the processor 11. For example, the storage unit in this embodiment may be realized by the main memory 13 or the auxiliary storage device 15.

[0029] Multiple processors may be connected to (combined with) one memory device, or a single processor 11 may be connected to it. Multiple memory devices may be connected to (combined with) one processor. If the information processing device 10 in this embodiment consists of at least one memory device and multiple processors connected to (combined with) this at least one memory device, it may include a configuration in which at least one of the multiple processors is connected to (combined with) at least one memory device. Furthermore, this configuration may be realized by memory devices and processors 11 included in multiple computers. In addition, it may include a configuration in which the memory device is integrated with the processor 11 (for example, a cache memory including an L1 cache and an L2 cache).

[0030] The network interface 17 in this embodiment is an interface for connecting to a network N wirelessly or via a wired connection. The network interface 17 can be any appropriate interface, such as one conforming to existing communication standards. Information may be exchanged between the information processing device 10 and external devices connected via the network N. The network N may be any of the following: WAN (Wide Area Network), LAN (Local Area Network), PAN (Personal Area Network), or a combination thereof, as long as information is exchanged between the information processing device 10 and external devices. An example of a WAN is the Internet; an example of a LAN is IEEE 802.11 or Ethernet (registered trademark); and an example of a PAN is Bluetooth (registered trademark) or NFC (Near Field Communication).

[0031] The device interface 18 in this embodiment is an interface such as a USB (Universal Serial Bus) that directly connects to an output device such as a display device, an input device, and an external device. The external device may be a storage device such as a memory device, network storage, or HDD. The external device may also be a device that has some of the functions of the components of the information processing device 10 in this embodiment. In other words, the information processing device 10 may transmit or receive some or all of the processing results of the external device.

[0032] <Functional Configuration (Information Processing Device)> Figure 3 shows an example of the functional blocks in the processor 11 and storage devices 13 and 15 of the information processing device 10 according to this embodiment. The processor 11 of the information processing device 10 includes, for example, an acquisition unit 101, a modification part identification unit 102, an analysis unit 103, a registration unit 104, an editing processing unit 105, and a first control unit 106 as functions realized by the processor 11. The functions realized by the acquisition unit 101, the modification part identification unit 102, the analysis unit 103, the registration unit 104, the editing processing unit 105, and the first control unit 106 are each stored as programs in, for example, the main memory 13 or the auxiliary storage device 15. The processor 11 can realize the functions related to the acquisition unit 101, the modification part identification unit 102, the analysis unit 103, the registration unit 104, the editing processing unit 105, and the first control unit 106 by reading and executing the programs stored in the main memory 13 or the auxiliary storage device 15.

[0033] In this embodiment, the acquisition unit 101 acquires DOCX format document data as a supported document, for example, via the network N. The acquisition unit 101 also acquires DOCX format document data stored in the first storage unit 107 as a supported document. Furthermore, the acquisition unit 101 acquires DOCX format document data stored in the storage unit of another device as a supported document, for example, via the network N.

[0034] The modification part identification unit 102 identifies, for example, the modified parts of the acquired support document. The method for identifying the modified parts may be one of the following, or a combination thereof: referring to the revision history information contained in the electronic file of the support document; obtaining the difference between the support document and the template document from which the support document was created or a previous version of the support document; or the user specifying the modified parts. These processes can be carried out using natural language processing, configured rules, machine learning models, etc.

[0035] In this embodiment, the analysis unit 103, for example, uses the document data of the document to be supported received from the acquisition unit 101 to perform a document review process on the modified portion of the document to be supported, based on a first criterion (for example, a checklist such as a playbook) stored in the first storage unit 107. That is, the analysis unit 103 in this embodiment performs a review process and generates document review information related to the document data.

[0036] The document review information generated by the document review process performed by the analysis unit 103 of this embodiment is based on information obtained by analyzing the document based on rules or standards applicable to the document (e.g., certain rules, laws, customs). For example, review information for a contract like the one in this embodiment may include: (1) whether the clauses contained in the document (the contract in this embodiment) are advantageous or disadvantageous to the user (the party); (2) advice on modifying or deleting clauses contained in the contract; (3) whether items that should normally be included in the contract are missing, and suggestions for items to be added; and (4) information on the importance and degree of recommendation of each item in the review results. The (document) review information exemplified in (1) to (4) above may be called contract review information when the supported document is a contract. By using the above information, the user can easily review the contract, including understanding the legal risks contained in the contract and suggesting modifications. It may also include formatting information such as whether the style of the text conforms to standards, and whether the prescribed terminology is used.

[0037] Performing a document review on the modified portion includes both performing a document review (including extraction and output of alert information) only on the modified portion, and performing a document review on a certain range including the modified portion. By including information about the surrounding text and context as part of the review process, it becomes possible to extract appropriate alert information. This can be achieved, for example, by inputting a prompt into the LLM that instructs it to match a certain range of text including the modified portion (e.g., a sentence or a paragraph) with a first criterion and output alert information based on the first criterion.

[0038] The analysis unit 103 may, for example, extract information contained in the document (document data) that is necessary for generating review information. Various techniques can be used to extract information from the document data. For example, it may perform a keyword search of the document data and extract characters, numbers, etc., contained near the keywords. Information can also be extracted by morphological analysis, syntactic analysis, semantic analysis, and contextual analysis. Furthermore, information may be extracted by natural language processing using machine learning. Prior to these processes, information necessary for document review, such as text data, may be extracted from the received document data, or processing such as OCR may be performed to extract text data from image data. Note that document review processing may be performed on the received document without extracting information necessary for generating document review information as described above. The information extraction process by the analysis unit 103 may be performed before the identification of the modified parts by the modified part identification unit 102. This may allow for appropriate identification of the modified parts.

[0039] The analysis unit 103 can employ various techniques for generating document review information. For example, it may generate document review information using a rule-based method with the extracted information. Alternatively, a trained model can be used that takes the extracted information as input and outputs the corresponding document review information. Furthermore, a combination of a rule-based method and a trained model can be employed. In addition, before generating the document review information, the document may be divided into units such as articles, paragraphs, or subparagraphs. These division units may correspond to document units.

[0040] If the document to be supported is divided into multiple document units, the document review process may be performed on the divided document unit containing the modified portion. In other words, a certain range containing the modified portion that is subject to the document review process may coincide with a document unit. In particular, if the document unit is at the level of articles, paragraphs, or subparagraphs, it is highly likely that the document unit has a cohesive meaning. Therefore, by making the document unit the target of alert information extraction, appropriate review information can be obtained. This can be achieved, for example, by inputting a prompt into the LLM that commands it to match the text of the divided document unit containing the modified portion with a first criterion and output alert information based on the first criterion.

[0041] The analysis unit 103 can also generate document review information tailored to the user's (party's) perspective by utilizing metadata associated with the document being supported. Furthermore, the analysis unit 103 can also generate document review information by referencing the contents of other documents (other versions of the document being supported, other contracts related to the document being supported, etc.). Metadata includes information attached to the document file itself, such as the document creator, as well as information added by the user, such as by being entered separately from the file.

[0042] The review information generated by the analysis unit 103 is displayed on the display unit of the terminal device T1, for example, in the form of an alert display.

[0043] Figure 4A shows an example of an alert display for review information provided to the user on terminal device T1. In Figure 4A, a support target document display unit 400a, which displays the support target document in an editable state, is provided on the left side of the screen, and an alert information display unit 401a, which displays alert information to draw the user's attention to the review information, is provided on the right side. Underlines or strikethroughs are provided for the modified parts of the support target document.

[0044] In this example, all three pieces of alert information displayed on the screen are labeled as "Meets fallback". That is to say, it is indicated that the corrected portion corresponds to content that is permitted to be corrected among the content of the playbook information used for the review process. Specifically, in the topmost alert information, it is indicated that the corrected portion related to "Assignment" corresponds to the content of "Fallback Position" that is included in the playbook information and displayed in the alert information.

[0045] In addition, the bottommost alert information displays "Analyzing", which indicates that the review process is being performed or is in the middle of being output.

[0046] By clicking the "Expand" button, the detailed description of each individual piece of alert information is displayed in the same manner as the topmost alert information.

[0047] In the present embodiment, by operating the scroll bar displayed on the screen or the like, the display sections of the support target document display unit 400a and the alert information display unit 401a can be scrolled, so that the continuation of the information displayed on the screen can be checked.

[0048] Here, the review process executed by the analysis unit 103 will be described.

[0049] The review process in this embodiment is performed based on a playbook. In other words, in this embodiment, the first criterion is the playbook. A playbook is a document containing information that an organization or individual (user) receiving a document viewing support service considers when reviewing a document to be supported, for example, and includes reference information such as the meaning and purpose of each provision, response policies, and response examples. In this embodiment, the various information contained in the playbook is also called "playbook information." The review process based on the playbook is performed using a trained model, such as an LLM, which can match the contents of the playbook with the contents of the document to be supported, for example, in a database of playbooks created by the recipient of the document viewing support service and stored in a storage device. In this specification, the criterion includes a determination criterion that determines whether or not to output review information (alert information), and the review information that is output when it is determined by the determination criterion that review information should be output.

[0050] In the review process of this embodiment, alert information is extracted from the modified portion of the document to be supported.

[0051] This allows for the display of alerts regarding revisions, which are considered more important than those regarding unrevised sections, thus supporting efficient document review.

[0052] Furthermore, using LLM for extracting alert information offers the following advantages. LLM can take time to analyze and output information when the amount of text or items related to alert extraction is large. Also, it tends to produce inappropriate results due to hallucination, etc. Therefore, it is preferable to not perform alert information extraction processing on parts of the supported document that have not been modified and do not require confirmation, and instead input only the modified parts into LLM, thereby appropriately outputting information to support the review. In addition, when using LLM provided as an external service, there may be costs associated with the amount of data input (for example, pay-per-use charges). As in this embodiment, by inputting only the modified parts of the supported document into LLM, a reduction in the costs associated with the review process can be expected.

[0053] Furthermore, the playbook information entered into the LLM can be the entirety of the playbook information contained in a single playbook, or only the portion related to the modification. This can be achieved by pre-analyzing the content of the modification (e.g., a summary of the clauses) using keywords or machine learning models, and then extracting the relevant parts of the playbook.

[0054] Furthermore, the alert extraction process may also be performed on parts that do not require confirmation and do not include the modified sections, and the output alert information may not be displayed. In this case, alert information for modified sections that are considered more important than those related to other sections will be displayed, which can support efficient document review.

[0055] In the review process of this embodiment, the criteria applied to the document to be supported may be set, for example, based on the type of contract of the document to be supported. Also, if the first criterion is a playbook, the playbook information may be set, for example, based on the description in the template document. Specifically, each part of the template document (e.g., document unit) can be associated with notes, alternative proposals, policies such as whether concessions or modifications are acceptable, etc.

[0056] Template documents are often composed of content that users deem appropriate based on their own perspectives, such as content that is most advantageous or low-risk to the user. Therefore, when a template document is modified, it is highly likely that the modified content will be less advantageous or higher-risk than the original template document.

[0057] By creating playbooks based on the contents of template documents, it becomes possible to meticulously and appropriately review support documents that require attention, which have been edited and modified from the template documents. If multiple template documents exist, it is preferable to set up a corresponding playbook and playbook information for each template document.

[0058] If playbook information is set for a template document, the playbook to be applied in the review process of the supported document may be automatically set by comparing the template document and the supported document. For example, a machine learning model such as LLM may be used to determine the degree of similarity or agreement of the wording, or the file name may be used to make a determination. Alternatively, the user may specify it manually. In this embodiment, in addition to the review process based on the playbook, which is the first criterion, a review process may be performed based on a second criterion different from the first criterion, for example, a highly general-purpose criterion prepared by the provider of the document viewing support service. This review process based on the second criterion may be performed using, for example, a database or a trained model created by the provider of the document viewing support service.

[0059] When a review process based on this second criterion is performed, the alert information shown in Figure 4A is generated based on a database created by the document viewing support service provider. By checking the alert information generated by the review process based on the second criterion, users can appropriately understand the general risks contained in the supported document.

[0060] Alert information based on playbook information and alert information based on the second criterion may be displayed simultaneously, or only one of them may be displayed by switching the display.

[0061] Furthermore, the first standard may be a highly versatile standard prepared by the provider of the document viewing support service.

[0062] Each alert may include information that identifies whether it is an alert based on playbook information or an alert based on a second criterion.

[0063] The registration unit 104 registers playbook information in the playbook database of the first storage unit 107 based on the information acquired by the acquisition unit 101. For example, the registration unit 104 registers playbook information entered by the user or generated / suggested by a trained model in the database of the first storage unit 107 and edits the database. Registration by the registration unit 104 may be performed by input from the screen of the terminal device T1, or by uploading data containing the necessary information (for example, a tabular file). This allows for the registration of playbook information for multiple playbooks at once, increasing user convenience. Furthermore, convenience is increased if the playbook information used by the user is stored in tabular format. Note that the first storage unit 107 in this embodiment is not limited to a storage device within the information processing device 10, but may be a storage device connected by a network within the system, a server outside the information processing system, etc.

[0064] Figure 4B shows another example of the alert display of review information provided to the user on terminal device T1. In Figure 4B, as in the example in Figure 4A, the support target document display unit 400b is displayed on the left and the alert information display unit 401b is displayed on the right.

[0065] In this example, two of the alerts displayed on the screen show "Violates non-negotiables." This indicates that the modified portion corresponds to content in the playbook information used in the review process that is unacceptable to modify. Specifically, the top alert indicates that the modified portion related to "Prior Disclosure" corresponds to the "Non-Negotiable Position" content included in the playbook information and displayed within the alert. As with the example in Figure 4A, clicking the "Expand" button in the area where the summary is displayed will show a detailed explanation of each alert, similar to the top alert.

[0066] The registration unit 104 registers playbook information in the database of the first storage unit 107, for example, based on information entered by the user via the GUI.

[0067] The playbook information of this embodiment includes various types of information.

[0068] The playbook information in this embodiment includes information about the modified sections for which alert information is set. This is expressed in words such as an overview or summary of the content of the modified section, such as "scope of disclosure of confidential information," or example clauses that apply. Although this information about the modified sections often does not perfectly match the wording of the modified sections in the actual contract, it can be used in document review processing by matching whether the information about the modified sections corresponds to the modified sections using keywords or machine learning models such as LLM.

[0069] In this embodiment, the playbook information includes information on the response policy regarding the modified portion.

[0070] Information regarding the response policy includes, for example, information on the policy regarding whether the content of the modified part is acceptable or unacceptable.

[0071] Information regarding the policy on modifications may be expressed using wording that includes the content of the modification and the course of action, such as: for example, in the case of acceptable modifications, "Changes to A are acceptable," or "If B is changed to C, it is acceptable if condition D is added"; or in the case of non-negotiable modifications, "Modifications are not permitted if E is F," or "No changes whatsoever are permitted for this content." It is also possible to set a policy such as "Review Required by the person in charge." Information regarding the response policy may be set by selecting from options prepared by the information processing system that indicate whether it is acceptable or not.

[0072] Furthermore, playbook information is not limited to being created by the user; it may also be proposed by the information processing system based on information prepared by the information processing system vendor according to the content and nature of the document, judgments and results from the user's past document review processes stored in the information processing system, and information appropriate for setting up playbook information collected from e-book databases, the internet, etc. The information processing system may also suggest corrections or improvements to the wording entered by the user.

[0073] Furthermore, the playbook information includes, for example, the following:

[0074] Fallback sample sentences are, for example, sample phrases to use when responding to a client about revisions or when suggesting further revisions.

[0075] The playbook information of this embodiment may further include guidance.

[0076] Guidance may include, for example, the criteria, purpose, interpretation, and precautions for a clause. It may also include tables and images as needed.

[0077] Furthermore, playbook metadata may be set to facilitate playbook management and document review. Examples of playbook metadata include the following:

[0078] - A unique ID to identify the playbook information. - Document type: This item specifies the type of document to which the playbook information applies. In this embodiment, the document to which the playbook information applies is a contract, and the document type (contract type) is set to a template document called "Non-Disclosure Agreement for sales department". The document type information may be, for example, information that refers to a specific template document, or it may be a type of contract. - Position: Indicates the user's position when reviewing the document. For example, if the contract type is "Non-Disclosure Agreement", this item indicates the position of the disclosing party, the receiving party, etc. By setting this item, appropriate alert information can be displayed for risks contained in the document that vary depending on the user's position. - Information that can identify the template document corresponding to the playbook, such as the ID of the template document, the name, the name of the department or person in charge of managing the template, the update date, etc.

[0079] In this embodiment, the analysis unit 103 performs a review process on the modified portion of the document to be supported by referring to playbook information stored in the database.

[0080] In other words, for example, if the clause concerning the "scope of disclosure of confidential information" is the amended part, the analysis unit 103 generates review information including playbook information about the "amendment concerning the scope of disclosure of confidential information" and displays it as alert information.

[0081] In this embodiment, the editing processing unit 105 converts the data format of the supported document (document data) received via the first control unit 106 and generates editing data. For example, the editing processing unit 105 converts the document data of a supported document, a "DOCX" format file received via the acquisition unit 101, from XML format to "JSON" format and generates editing data.

[0082] Furthermore, the editing processing unit 105 of this embodiment performs editing processing on the editing data in response to editing instructions received from the terminal device T1 via the acquisition unit 101. Here, the editing processing performed by the editing processing unit 105 includes adding, deleting, and changing information contained in the editing data. More specifically, this includes adding, modifying, and deleting words, sentences, and text formatting (e.g., font, text color, text size, text weight, underline, strikethrough, etc.) to the editing data.

[0083] Furthermore, the editing process performed by the editing processing unit 105 may include formatting editing, word input assistance, text suggestion, issue extraction, and difference information extraction.

[0084] Here, format editing processing refers to, for example, the process of correcting the format of the editing data as a document. For example, format editing processing preferably automatically corrects indentation, bullet point alignment, right / left alignment, and misalignment of clause numbers by the system. The editing processing unit 105, for example, refers to the format information of the support document corresponding to the editing data and executes format editing processing on the editing data based on this information.

[0085] It is preferable that the formatting settings for a document are stored in advance on the information processing device 10 or terminal, etc., based on the type of document to be supported, such as a contract. For example, in the case of legal documents such as contracts or internal regulations, the clauses may be arranged in the order of articles, paragraphs, and subparagraphs, with the clause numbers decreasing in ascending order, and the indentation may increase in the order of articles, paragraphs, and subparagraphs. The editing processing device 105 may also perform formatting editing processing according to the type based on the metadata of the document to be supported. This formatting setting may be created by the user, provided as part of the document viewing support service, or customized by the user from the provided setting. Alternatively, it may be created by interpreting the formatting setting of the document to be supported.

[0086] The word input assistance process in this embodiment is a process that assists input by, for example, presenting candidate words or phrases to be entered next when editing document data. This supports the user's input operations during editing and reduces the workload. For example, frequently occurring words can be presented as input candidates depending on the type of document being supported. Alternatively, the system may analyze the documents being supported stored by the user or the organization to which the user belongs and present commonly used words. The user may also set the words or phrases. The editing processing unit 105 can determine candidate words according to the type of document based on the metadata of the document being supported. The words presented in this input assistance process are displayed, for example, in a predetermined window on the screen (within the document editing window).

[0087] The text suggestion process in this embodiment is, for example, a process that displays suggested text during editing. For example, if the document is a contract, it may retrieve a revised text example based on at least a portion of the revisions to the contract from a database containing examples of draft clauses and present it as a suggested text. This supports the user's input operations during editing and reduces their workload.

[0088] The editing processing unit 105 can, as part of its text suggestion process, present, for example, text candidates included in playbook information. It may also present text candidates generated by an in-system or external document generation service using LLM, etc. It may also present frequently occurring text candidates depending on the type of document being supported. Furthermore, it may present text candidates that are more frequently used than other texts or that have received high evaluations from the user or users including the user, among the texts contained in documents stored by the user or the user's organization. If multiple text candidates exist, it can further select at least one text candidate based on the adoption rate of clauses in past document viewing support services (the percentage of proposed clauses selected by users) and the frequency of adoption, and present the candidate with the highest adoption rate to the user preferentially. The presented text candidates can also be presented, for example, as document review information or as other information.

[0089] The editing processing unit 105 may automatically reflect draft candidates in the document to be supported, or it may edit the editing data in response to editing instructions, such as instructions via drag-and-drop. The proposed drafts may be displayed with their relative positions aligned with the corresponding parts of the document. For example, draft candidates may be displayed close to the corresponding parts of the document. Such a display allows users to compare draft candidates and the corresponding parts of the document with minimal eye movement, thereby reducing the burden on the user. In addition, parts of the editing data may be replaced with the presented draft candidates according to user instructions or automatically. In this case, it is preferable to add a revision history to the editing data so that it is clear what edits have been made.

[0090] The processing for displaying the issues in this embodiment involves, for example, presenting the user with issues related to the document being supported (e.g., document review information, comment information, chat information, revision history for the editing data, comments on document data created by the functions of the document creation software, etc.). The editing processing unit 105 in this embodiment presents the extracted issues to the user in a predetermined display format. Preferably, the issues are displayed in a list in an area that does not overlap with the part of the document that the issues concern, for example, in a document review area displayed alongside the editing area, or in an independent (individual) window such as a pop-up window.

[0091] Furthermore, by combining the processing of displaying the issues and the processing of suggesting alternative text, the issues and alternative text options that can resolve the issues may be suggested. Specifically, LLM may be used to suggest alternative text options by processing the text from the part of the document from which the alert information has been extracted into text that can resolve the issues. In this case, it is preferable that the issues and alternative text options are displayed in a positional correspondence with the relevant part of the document. For example, it is preferable that the issues and alternative text options are displayed close to the corresponding issues. Such a display allows users to compare the issues, alternative text options and the relevant part of the document with minimal eye movement, thereby reducing the burden on the user.

[0092] Furthermore, the document review area and pop-up windows displaying comments and suggested drafts can be collapsed, minimized, and resized. This allows users to arrange the limited document display screen in a way that is easy to view, for example, by aligning it with the highlighted portion of the document. The spacing between clauses in the editing area may also be widened based on the height of the comments or suggested drafts. This type of display makes the correspondence between the comments, suggested drafts, and the highlighted portion of the document clearer and easier to understand.

[0093] The difference information extraction process in this embodiment presents the user with difference information between two document data sets, for example, the document data to be supported before modification and the document data after modification. The editing processing unit 105 in this embodiment may, for example, extract difference information from the document data to be supported before modification and the document data after modification, and present it to the user in a predetermined display format, such as displaying the difference and the other parts (i.e., the same parts) differently, for example, by highlighting them. This difference information extraction process may be the same as the process performed when identifying the modified part. Alternatively, the modified part identification information obtained in the modified part identification process may be used as difference information, or the information obtained in the difference information extraction process may be used as modified part identification information.

[0094] The editing processing unit 105 of this embodiment performs the various editing processes described above in response to editing instructions received from the terminal device T1. Furthermore, the editing processing unit 105 of this embodiment may convert the data format of the supported document data or translation data in response to a save instruction received from the terminal device T1, and generate document data that will be the next version of the supported document. For example, the editing processing unit 105 can generate document data in DOCX format that will be the next version of the supported document using supported document data in JSON format in response to a download or save instruction received from the terminal device T1.

[0095] In this embodiment, the first control unit 106 communicates with the terminal device T1 via a network. Specifically, for example, the first control unit 106 receives various instructions from the terminal device T1, such as document data, document review instructions, editing instructions, translation instructions, save instructions, and output instructions (e.g., download instructions). The first control unit 106 also transmits editing screen information (including information for executing editing processing on the terminal device T1 side) to the terminal device T1. The first control unit 106 receives information from the terminal device T1, including the document data after editing processing.

[0096] The first control unit 106 of this embodiment generates layout information during document editing. Layout information in document editing refers to, for example, information for displaying editing data in a predetermined format on the browser screen of the terminal device T1. The first control unit 106 of this embodiment may also generate layout information during document review. Layout information in document review refers to, for example, information for displaying the document review information generated by the document review process and the supported document in a predetermined format on the browser screen of the terminal device T1.

[0097] Layout information may map the document review information (including translated information) generated by the document review process to the document being supported. This mapping means, for example, mapping the displayed positions to facilitate user viewing, specifically, mapping positions on the screen, for example, at the clause, paragraph, or sentence level. Mapping positions includes, for example, displaying corresponding parts closer in height or width than other parts on the screen, more preferably side-by-side. Furthermore, when additional information is displayed, the layout information may include information for mapping the additional information, document review information, and editing data to each other. It may also include information that, when the display of alert information is selected, for example by clicking, the screen scrolls to display the modified portion to which the alert information is attached. It may also include information that, when the display of the modified portion is selected, for example by clicking, the attached alert information is displayed.

[0098] The generation of layout information can be done by any method. For example, the degree of matching of words and phrases included in the document review information may be scored for each position in the editing data (for example, each position at the clause level or paragraph level) using a search function, and the layout information may be generated so that the document review information is displayed corresponding to the position with the highest score. Alternatively, for example, coordinates can be assigned to each clause or paragraph in the editing data, and these coordinates can be associated with the corresponding document review information to establish a correspondence between the display position of clauses, paragraphs and document review information.

[0099] In this embodiment, the first control unit 106 transmits various types of information, such as document review information generated in the analysis unit 103, editing data generated in the editing processing unit 105, translation data, and layout information generated in the first control unit 106, in response to a request from the terminal device T1.

[0100] The first storage unit 107 stores, for example, document data uploaded from terminal device T1, new versions of document data generated from editing data through document editing processing, DOCX format supported documents, and additional information associated with each document data.

[0101] Furthermore, the first storage unit 107 stores a first database and a second database. Here, the first database is a database in which playbook information is registered. In this embodiment, the playbook information registered in the first database can be edited by an editing process performed by a user with editing privileges. The second database is a database in which multiple alert information is registered that draws the user's attention to the content of document units that constitute a document based on a second criterion. Multiple playbooks are stored in these databases.

[0102] Thus, in one embodiment of the present disclosure, the information processing device 10 can perform a document review on the modified portion (i.e., the portion modified from the template) of a supported document created by modifying a template of a document such as a contract. For example, the template and the supported document are divided into multiple document units such as articles, paragraphs, and clauses. The information processing device 10 can perform a document review process (for example, outputting whether the modification from the template is acceptable or unacceptable) using playbook information corresponding to the document unit containing the modified portion of each template.

[0103] (Terminal device T1) In this embodiment, terminal device T1 transmits and receives information related to document display and document editing with the information processing device 10, for example, in a document viewing support service (document viewing support system).

[0104] <Hardware Configuration (Terminal Device)> Figure 5 is a block diagram showing an example of the hardware configuration of a terminal device T1 according to this embodiment. As shown in Figure 5, the terminal device T1 is a computer and, as an example, includes a processor 21, main memory 23, auxiliary memory 25, network interface 27, device interface 28, a bus 29 connecting each component, a display device 30 connected to the device interface 28, and an input device 31 connected to the device interface 28. Note that the specific hardware configurations of the processor 21, main memory 23, auxiliary memory 25, network interface 27, device interface 28, and bus 29 are the same as those of the processor 11, main memory 13, auxiliary memory 15, network interface 17, device interface 18, and bus 19, respectively, so their explanation is omitted.

[0105] The display device 30 is a display device such as an LCD (Liquid Crystal Display), CRT (Cathode Ray Tube), PDP (Plasma Display Panel), or organic EL (Electroluminescence) panel. Based on instructions from the processor 21, the display device 30 displays, for example, a document review screen or a document editing screen, which will be described later.

[0106] The input device 31 is a device such as a keyboard, mouse, or touch panel. The user can use the input device 31 to input various editing instructions, for example, in document editing processing. The instructions input from the input device 31 are sent to the processor 21.

[0107] The terminal device T1 may also be equipped with other external devices (for example, a camera, microphone, motion capture device, various sensors, etc.) connected via the device interface 18 for manipulating and editing document data.

[0108] <Functional Configuration (Terminal Device)> Figure 6 shows an example of a functional block in the processor 21 and storage devices 23 and 25 of the terminal device of this embodiment. The processor 21 of the terminal device T1 includes, for example, a second control unit 201 and a browser unit 202 as functions realized by the processor 21. The functions realized by the second control unit 201 and the browser unit 202 are stored as programs in, for example, the main memory 23 or the auxiliary storage device 25. The processor 21 can realize the functions related to the second control unit 201 and the browser unit 202 by reading and executing the programs stored in the main memory 23 or the auxiliary storage device 25.

[0109] The second control unit 201, for example, starts the browser unit 202 in response to instructions from the user.

[0110] The browser unit 202 is, for example, a general web browser and executes processing according to HTML (Hyper Text Markup Language) data and scripts (for example, JavaScript®). When the browser unit 202 is started, it sends an initial access URL to the information processing device 10 and accesses it.

[0111] In this embodiment, the browser unit 202 displays data, including text data and image data received from the information processing device 10, on the display device 30 of the terminal device T1 in a predetermined format. More specifically, in document review processing, the browser unit 202 displays document review information, document data, additional information, etc., received from the information processing device 10 on the display device 30 of the terminal device T1 in a predetermined format based on layout information. Furthermore, in document editing processing, the browser unit 202 of this embodiment displays, for example, document review information, editing data, additional information, etc., received from the information processing device 10 on the display device 30 of the terminal device T1 in a predetermined format based on layout information.

[0112] Furthermore, the browser unit 202 transmits, for example, a start command from the user to the information processing device 10. For example, the browser unit 202 receives playbook information registration screen information (including information for executing the registration process) from the first control unit 106 of the information processing device 10 and executes the registration process, etc. The browser unit 202 receives, for example, the supported document for which the registration process has been executed and stores it in the second storage unit 203 in the storage device of the terminal device T1.

[0113] The second storage unit 203 stores, for example, the document data to be supported before editing and the document data to be supported after editing in DOCX format acquired by the browser unit 202. The second storage unit 203 may also store, for example, playbook information.

[0114] [First Example of Review Processing] Next, a first example of review processing implemented by the document viewing support system S will be described. In this embodiment, playbook information associated with a template document exists, and for a support target document whose contents have been modified from the template document, review processing is performed using the associated playbook information.

[0115] Figure 7 is a flowchart showing the flow of document viewing support processing (information processing) including review processing according to the first example. As shown in Figure 7, first, the second control unit 201 of the terminal device T1 starts the browser unit 202 in response to instructions from the user, for example. The browser unit 202 of the terminal device T1 transmits login information such as the initial access URL and registration ID to the information processing device 10 (step S11).

[0116] The first control unit 106 of the information processing device 10 performs authentication processing in response to login information from terminal device T1, for example. The first control unit 106 also transmits document viewing support screen information generated in the first control unit 106 (step S12).

[0117] The browser unit 202 of the terminal device T1 displays a document viewing support screen on the display device 30 based on the document viewing support screen information received from the information processing device 10 (step S13).

[0118] The browser unit 202 of the terminal device T1, for example, responds to a document upload instruction from the user by sending the supported document data to the information processing device 10 (step S14).

[0119] The acquisition unit 101 of the information processing device 10 stores document data acquired (received) from terminal device T1 in the first storage unit 107. The first control unit 106 of the information processing device 10 generates document preview screen information based on the document data acquired by the acquisition unit 101. The first control unit 106 of the information processing device 10 transmits the generated document preview screen information to terminal device T1 (step S15). Note that the document upload instruction and the saving of document data to the first storage unit 107 may be performed, for example, in advance of the editing process, by another user in another department or / or another terminal requesting the legal department to review a contract. The document data may be subjected to determination of the document type of the document to be supported and analysis of document units.

[0120] The browser unit 202 of the terminal device T1 displays a document preview screen based on the document preview screen information received from the information processing device 10 (step S16). At this stage, since the document review process has not yet been completed, only the document to be supported is displayed on the document preview screen. The document viewing support system S of this embodiment can be used not only for reviewing and editing documents but also for viewing documents by executing step S16. The display of the document preview screen can be omitted by receiving a document review instruction or performing the document review process, as described later, before displaying the document preview screen. If the parts to be modified have already been identified before the document review process, the modified parts may be displayed in a different manner from the other parts. For example, the modified parts may be highlighted, underlined, bolded, etc.

[0121] The browser unit 202 of the terminal device T1 transmits, for example, a document review instruction entered by the user to the information processing device 10 (step S17).

[0122] The modification portion identification unit 102 of the information processing device 10 identifies the modified portion of the supported document based on, for example, the modification history information contained in the supported document data, in response to a document review instruction from the terminal device T1 (step S18).

[0123] Next, the analysis unit 103 of the information processing device 10 performs a review process on the identified modification portion and generates review information (step S19).

[0124] Figure 8 is a flowchart showing the flow of the review process related to the first example.

[0125] For example, the analysis unit 103 of the information processing device 10 refers to a database in which multiple playbooks are pre-stored and determines whether a template document that served as the basis for the document to be supported and a playbook associated with that template document exist (step S191a). If no playbook is found, the process terminates.

[0126] If it is determined that playbook information associated with a template document exists, the analysis unit 103 of the information processing device 10 determines whether the playbook contains playbook information associated with the modified portion (Yes in step S192a), and generates review information associated with the modified portion based on the contents of the playbook information and its metadata (playbook metadata) (step S193a).

[0127] The analysis unit 103 of the information processing device 10 determines whether the following modification exists (step S194a). If it determines that a clause containing the following modification exists (Yes in step S194a), the analysis unit 103 of the information processing device 10 repeatedly executes the processes in steps S192a to S194a to update the review information. On the other hand, if it determines that the following modification does not exist (No in step S194a), the analysis unit 103 of the information processing device 10 determines the content of the review information (step S195a).

[0128] Returning to Figure 7, the first control unit 106 of the information processing device 10 generates layout information for displaying the review information in association with the document to be supported. The first control unit 106 of the information processing device 10 transmits the document to be supported data, information identifying the modified portion in the document to be supported data, the review information, and the layout information generated by the first control unit 106 to the terminal device T1 (step S20).

[0129] The browser unit 202 of the terminal device T1 displays a document review screen based on the information received from the information processing device 10 (step S21). That is, for example, alert information based on the playbook is displayed as a document review screen along with information identifying the modified portion, for example, in a form aligned with the corresponding clause. It is preferable that the modified portion associated with alert information is displayed in a different manner from the unmodified portion or the modified portion that does not have alert information. For example, it may be highlighted, underlined, or bolded.

[0130] Furthermore, the display order of playbook information on the document review screen can be controlled based on, for example, the importance level and update date included in the playbook metadata. Additionally, playbook information can be toggled on or off based on at least one of the following: the type of document being supported, the importance level of the clauses, or the user's position.

[0131] Furthermore, the review information may include status information indicating whether the support for the document has been completed or not, based on the alert information in the playbook. This status can also be changed in response to user input on the document review screen.

[0132] The information processing device 10 according to this embodiment comprises a first database, an acquisition unit 101, a modification portion identification unit 102, an analysis unit 103, and a first control unit 106. The first database is, for example, a playbook information database in which multiple playbook information that draws the user's attention to the content of document units constituting a document based on a first criterion (for example, a unique (individual) criterion of the party receiving the document viewing support service) is registered for each type of document unit. The acquisition unit 101 acquires the document to be supported. The modification portion identification unit 102 identifies the modified portion of the document to be supported. The analysis unit 103 extracts at least one playbook information from the playbook information database, performs a review process on the modified portion of the document to be supported, and generates review information including the extracted alert information and the extracted playbook information. The first control unit 106 displays the review information and the corresponding document unit to be supported on the display device 30 of the terminal device T1.

[0133] Therefore, users can perform a review process on the modified parts of the document being supported and check the results. As a result, it is possible to provide support technology that enables more efficient review processing of documents.

[0134] The step of identifying the parts to be corrected may, but is not limited to, being performed after the review process instructions. For example, it can be performed at an appropriate time, such as immediately after the document to be supported is acquired, following various analyses by the analysis unit 103, or before the document preview is performed.

[0135] (Modification 1) In the above-described embodiment, some or all of the devices may be composed of hardware, or they may be composed of information processing by software (programs) executed by a CPU or GPU, etc. Also, in the above embodiment, if multiple storage devices store data, each of the multiple storage devices may store only a part of the data, or it may store the entire data.

[0136] Some or all of the processing in the above-described embodiment may be performed by creating an appropriate prompt in the information processing device 10, inputting the prompt to an LLM provided in an internal or external server of the information processing system S, and receiving the output.

[0137] Although embodiments of this disclosure have been described in detail above, these embodiments are presented as examples only and are not intended to limit the user to individual embodiments. Each embodiment can be modified in various ways, including additions, changes, substitutions, partial deletions, and combinations, without departing from the technical spirit of the present invention. These embodiments and their variations are included within the scope of the invention described in the claims and its equivalents.

[0138] This international application claims priority under U.S. application 63 / 790,345, filed on 17 April 2025, and the entire contents of 63 / 790,345 are incorporated herein by reference.

Claims

1. An information processing method comprising a computer that identifies a modified portion in a document to be supported, and extracts at least one alert information based on a first criterion having one or more alert information to draw the user's attention to the modified portion of the document to be supported.

2. The information processing method according to claim 1, wherein the first criterion is a playbook set by the user.

3. The information processing method according to claim 1, wherein the document to be supported is a modified version of the template document used by the user.

4. The information processing method according to claim 3, wherein the multiple alert information is set based on the description in the template document.

5. The information processing method according to claim 1, wherein the identification of the modified portion is performed based on either revision history information contained in the electronic file of the supported document or difference information which is the difference between the supported document and a previous version of the supported document.

6. The information processing method according to claim 1, wherein the document to be supported is divided into multiple document units, and the extraction of the alert information is performed on the divided document unit including the modified portion.

7. The information processing method according to claim 1, wherein the document to be supported is a contract.

8. The information processing method according to claim 1, wherein the alert information includes at least information on whether the modified portion is acceptable or unacceptable.

9. An information processing device comprising at least one processor, which performs an information processing method for identifying a modified portion in a supported document, and for extracting at least one alert information for the modified portion of the supported document based on a first criterion having one or more alert information for drawing the user's attention.

10. An information processing system comprising at least one processor, which performs an information processing method for identifying a modified portion in a supported document, and for extracting at least one alert information for the modified portion of the supported document based on a first criterion having one or more alert information for drawing the user's attention.

Images