Information processing device, information processing method, and program

Abstract

To provide appropriate answers to user questions in a short amount of time. [Solution] An information processing device comprising: an acquisition unit that acquires content that is the subject of a question from a user; a division unit that divides the content into a plurality of regions according to the context; a first generation unit that generates a summary for each of the plurality of regions; a second generation unit that generates index information linking the summary, the category of the content from which the summary is to be compiled, and a label indicating the content from which the summary is to be compiled; a reception unit that receives a question from the user that is associated with the category and the label; a filtering unit that narrows down the index information based on the category and the label associated with the question; an extraction unit that extracts index information related to the question from the filtered index information; and a third generation unit that generates an answer to the question based on the extracted index information.

Description

Technical Field

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

Background Art

[0002] Conventionally, techniques for making it easier to obtain the answers that users want are known.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Patent Document 2

Summary of the Invention

Problems to be Solved by the Invention

[0004] However, in the conventional technology, there is room for further improvement in providing appropriate answers to user questions in a short time.

[0005] The present application has been made in view of the above, and an object thereof is to provide an appropriate answer to a user's question in a short time.

Means for Solving the Problems

[0006] The information processing device according to the present invention comprises: an acquisition unit that acquires content that is the subject of a question from a user; a division unit that divides the content into a plurality of regions according to the context; a first generation unit that generates a summary for each of the plurality of regions; a second generation unit that generates index information linking the summary, the category of the content from which the summary is to be compiled, and a label indicating the content from which the summary is to be compiled; a reception unit that receives a question from the user that is associated with the category and the label; a filtering unit that narrows down the index information based on the category and the label associated with the question; an extraction unit that extracts index information related to the question from the filtered index information; and a third generation unit that generates an answer to the question based on the extracted index information. [Effects of the Invention]

[0007] According to one embodiment, the system has the effect of providing appropriate answers to user questions in a short amount of time. [Brief explanation of the drawing]

[0008] [Figure 1] Figure 1 is an explanatory diagram illustrating the overview of the information processing according to the embodiment. [Figure 2] Figure 2 shows an example of the configuration of an information processing system according to the embodiment. [Figure 3] Figure 3 shows an example of information processing according to the embodiment (Embodiment 1). [Figure 4] Figure 4 shows an example of information processing according to the embodiment (Embodiment 2). [Figure 5] Figure 5 shows an example of content managed by a company according to this embodiment. [Figure 6] Figure 6 shows an example of the home screen of the chatbot according to this embodiment. [Figure 7] Figure 7 shows an example of a chatbot window according to the embodiment. [Figure 8]Figure 8 shows an example of an internal inquiry form according to this embodiment. [Figure 9] Figure 9 shows an example of the configuration of a terminal device according to an embodiment. [Figure 10] Figure 10 shows an example of the configuration of an information processing device according to an embodiment. [Figure 11] Figure 11 shows an example of an index information storage unit according to an embodiment. [Figure 12] Figure 12 shows an example of a response evaluation information storage unit according to an embodiment. [Figure 13] Figure 13 is a flowchart showing an example of information processing according to the embodiment (Embodiment 1). [Figure 14] Figure 14 is a flowchart showing an example of information processing according to the embodiment (Embodiment 2). [Figure 15] Figure 15 is a hardware configuration diagram showing an example of a computer that implements the functions of an information processing device. [Modes for carrying out the invention]

[0009] The following describes in detail, with reference to the drawings, the embodiments for implementing the information processing apparatus, information processing method, and program according to the present application (hereinafter referred to as "embodiments"). Note that these embodiments do not limit the information processing apparatus, information processing method, and program according to the present application. Furthermore, the same parts are denoted by the same reference numerals in each of the following embodiments, and redundant descriptions are omitted.

[0010] (Embodiment) [1. Overview of Information Processing] Using FIG. 1, the outline of the information processing according to the embodiment will be described. FIG. 1 is an explanatory diagram for explaining the outline of the information processing according to the embodiment. As shown in FIG. 1, the outline R includes outlines R1 to R6. The outline R1 is the acquisition of contents such as regulations, web contents, store information (sales history, etc.). Using the sentences in this content, answers to the questions of the user U are generated. The outline R2 is an index builder, and the content acquired in the outline R1 is indexed. The outline R3 is the acquisition of the indexed information (index information) indexed in the outline R2. The outline R4 is a dispatcher, and tasks based on the operations of the user U are executed. For example, a task for answering the question received from the user U is executed. The outline R5 is a prompt builder, and a prompt is generated using the index information acquired in the outline R3 and the question received in the outline R4. For example, the index information corresponding to the question received in the outline R4 is extracted and a prompt is generated from the index information. The outline R6 is a generative AI model such as GPT, and an answer to the question of the user U is generated based on the prompt generated in the outline R5. The answer generated in the outline R6 is provided to the user U via the dispatcher.

[0011] [2. Configuration of Information Processing System] The information processing system 1 shown in FIG. 2 will be described. As shown in FIG. 2, the information processing system 1 includes a terminal device 10 and an information processing device 100. The terminal device 10 and the information processing device 100 are communicably connected by wire or wirelessly via a predetermined communication network (network N). FIG. 2 is a diagram showing a configuration example of the information processing system 1 according to the embodiment.

[0012] The terminal device 10 is an information processing device used by a user who makes inquiries. The user is, for example, a user who makes inquiries regarding the contents of regulations, laws, rules, manuals, and the like. For example, a user who makes an inquiry such as "What is the upper limit of the accommodation fee in the company's accommodation regulations for a business trip?" The user inputs an inquiry, for example, using a chat tool capable of text-based comment exchanges. The terminal device 10 may be any device as long as it can realize the processing in the embodiment. Also, the terminal device 10 may be a device such as a smartphone, a tablet terminal, a notebook PC, a desktop PC, a mobile phone, or a PDA. FIG. 4 described later shows the case where the terminal device 10 is a smartphone.

[0013] The terminal device 10 is, for example, a smart device such as a smartphone or a tablet, and is a portable terminal device capable of communicating with an arbitrary server device via a wireless communication network such as 4G to 5G (Generation) or LTE (Long Term Evolution). Also, the terminal device 10 has a screen such as a liquid crystal display, and has a screen having a touch panel function, and may accept various operations on display data such as content, such as tap operations, slide operations, and scroll operations, by a finger or a stylus from the user. FIG. 4 described later shows the case where the terminal device 10 is used by the user U1.

[0014] The information processing device 100 is an information processing device intended to provide appropriate answers to user questions, and can be any device as long as it can realize the processing in the embodiment. The information processing device 100 performs two main types of processing. The first is processing before receiving a question from the user, which is processing to enable the provision of appropriate answers when a question is received from the user. The second is processing when a question is received from the user, which is processing to generate an answer to the user's question using the information from the result of the first processing. Specifically, in the first processing, the information processing device 100 generates information (hereinafter referred to as "index information" as appropriate) that summarizes content (such as regulations, laws, rules, and manuals) by chapter, article (clause), sentence, or main idea (a main idea with a predetermined meaning) for each predetermined unit (hereinafter referred to as "domain" in the embodiment). The generated index information is then used in the second processing. Furthermore, in the second processing, when the information processing device 100 receives a question from the user, it inputs the index information generated in the first processing into the generating AI to generate an answer to that question.

[0015] Here, the index information according to the embodiment will be described. The Lama Index and the like are used to generate the index information according to the embodiment. The Lama Index is a tool for generating context from a large amount of user data, and is a tool that performs data retrieval by inputting a question received from the user and the individual user data. In other words, the Lama Index accesses information sources in a database and assists in the generation of prompts. The index information is structured to organize the data taken from such information sources and make it easier to search. By using such index information, the information processing device 100 can efficiently access private information when generating prompts and generate contexts that include private information. When the information processing device 100 receives a natural language question from the user, it uses the Lama Index and the like to search for index information that matches the question from the user and inputs it into the generation AI.

[0016] Generative AI is a text generation model that has been trained on data publicly available on the internet, for example. A GPT (Generative Pre-trained Transformer) model is one such model that generates answers to questions. Such models generate answers to questions from publicly available data on the internet. Because new data is published on the internet daily, the model's answers are updated daily. This means that even if the same question is asked, the data the model references when generating the answer will differ depending on when the question is asked, resulting in different answers. Furthermore, because the model can provide detailed answers to questions across a wide range of fields, users ask the model a diverse variety of questions.

[0017] Although Figure 2 shows a case where the terminal device 10 and the information processing device 100 are separate devices, the terminal device 10 and the information processing device 100 may be integrated into a single unit.

[0018] [3. An example of information processing] Conventional technologies have been developed to make it easier for users to obtain the answers they want. For example, a technology is known for providing an internal chatbot system using internal company data (Patent Document 1 above). Also, for example, a technology is known for dividing regulations, manuals, etc., into predetermined units such as chapters and paragraphs and tagging them (Patent Document 2 above). However, conventional technologies do not, for example, appropriately narrow down the information used to generate the answer before generating the answer, so there was room for further improvement in providing appropriate answers to user questions in a short time. This application has been made in view of the above, and aims to provide appropriate answers to user questions in a short time by narrowing down the information used to generate the answer (index information) based on categories and labels before generating the answer.

[0019] Below, two information processing steps are described as part of the information processing of the information processing system 1 according to the embodiment. As described above, the first information processing step (Embodiment 1) is for generating index information, and the second information processing step (Embodiment 2) is for generating answers to user questions. By combining the two information processing steps, it becomes possible to provide appropriate answers to user questions. In the first information processing step (Indexing Stage), the information processing device 100 organizes and indexes the data acquired from the data source, and in the second information processing step (Querying Stage), it analyzes the question entered by the user, extracts relevant information from the index, and generates the final response. By using this mechanism, Lama Index can efficiently manage large amounts of data, and users can quickly obtain the information they need. This is particularly useful when a lot of information is scattered or when it is necessary to quickly retrieve specific information, and it significantly improves the efficiency of generating prompts to the generating AI, enabling the acquisition of better results.

[0020] (Embodiment 1: Information processing for generating index information) Figure 3 shows an example of information processing in the information processing system 1 according to Embodiment 1. The information processing device 100 acquires content (Content A) that is the subject of a question from the user (step S101). The content in this embodiment is an example of terms and conditions, laws, rules, or manuals, but it can be any kind of content and is not particularly limited. In the embodiment described later, an example is given where the content is "travel expense regulations", but the content may also be "service terms of use" which define the rights and obligations of the service provider and the compliance matters of the user between the service provider and the user. The content in this embodiment may be converted into a document object such as a PDF. That is, information acquired from different databases may be converted into a document object. The document object converted in this way may be converted into a more user-friendly node and indexed for efficient data access.

[0021] The information processing device 100 divides the acquired content into multiple areas (area A1, area A2, area A3, ...) (step S102). For example, the information processing device 100 divides the content into areas such as chapters, articles, sentences, and main points. The size of the divided areas (scope of area division) may be determined from the context of the content. In other words, the information processing device 100 may divide the content into areas according to the context of the content.

[0022] The information processing device 100 generates summaries (summary A1, summary A2, ...) for each divided area (step S103). Here, the generated summaries are, for example, "Business trip expenses are capped at 5,000 yen per night." These generated summaries are included in the index information.

[0023] The information processing device 100 generates index information (index P1, index P2, ...) by associating the generated summary with the category of the content from which the summary is taken and a label indicating the content from which the summary is taken (which may also be a label indicating the entire text) (step S104). The category (category B1, B2, ...) is information indicating, for example, which department created the content from which the summary is taken. For example, the category may be "Legal Department," "Human Resources Department," "General Affairs Department," etc. The label (label C1, label C2, ...) is a label indicating, for example, what kind of regulations, laws, rules, or manuals they are. For example, in the case of content related to regulations, the label may be "Travel Expense Regulations for Business Trips," etc. In addition, the information processing device 100 may generate index information by associating, for example, the summary with the category and label and a label indicating the scope within the content from which the summary is taken (hereinafter referred to as "scope label" as appropriate). Scope labels (scope label D1, scope label D2, ...) are labels that indicate, for example, which chapter, article, or sentence within the content the summary refers to. For example, a scope label might be "Article 5". To give a concrete example, the information processing device 100 generates index information by linking a summary such as "Business trip expenses are capped at 5,000 yen per night", a category such as "Legal Department", a label such as "Business Trip Expense Regulations", and a scope label such as "Article 5". Labels may also be information that indicates the parent-child relationship with content, and scope labels may also be information that indicates the parent-child relationship with content or labels.

[0024] Thus, in Embodiment 1, the information processing device 100 divides the content into areas, generates a summary for each divided area, assigns categories and labels to the generated summaries, and uses them as index information. For example, the information processing device 100 assigns a category related to a department, such as "Legal Department," a label related to regulations, such as "Business Trip Expense Regulations," and a label related to an article, such as "Article 5," to a summary such as "Business trip expenses are capped at 5,000 yen per night."

[0025] (Embodiment 2: Information processing for generating answers to user questions) Figure 4 shows an example of information processing in the information processing system 1 according to Embodiment 2. The information processing device 100 receives a question Q1 from user U1 (step S201). In the following embodiment, we will explain using the example of when the information processing device 100 obtains information regarding question Q1 from user U1, such as "What is the upper limit for accommodation expenses in the company's accommodation regulations?". Note that question Q1 is associated with a category and a label based on input from user U1. In Embodiment 2, when a user's question is received, appropriate information is searched from the index information, relevant nodes are retrieved, the necessary information is selected and becomes part of the query to the generating AI. As a result, it becomes possible to efficiently extract highly relevant information from a large amount of data and generate a natural language response to address the user's question.

[0026] The information processing device 100 narrows down the index information based on the category and label associated with question Q1 (step S202). For example, if question Q1 received from user U1 is associated with the category "Legal Department" and the label "Business Trip Expense Regulations", the information processing device 100 narrows down the index information based on the conditions that the category is "Legal Department" and the label is "Business Trip Expense Regulations".

[0027] The information processing device 100 extracts index information related to user U1's question Q1 from the narrowed-down index information (i.e., index information that summarizes predetermined content for each region) (step S203). For example, the information processing device 100 may generate a first vector by vectorizing user U1's question Q1 and a second vector by vectorizing the summary associated with the index information, and determine whether or not the index information is related to user U1's question Q1 by comparing the first vector and the second vector. For example, the information processing device 100 may use a BERT (Bidirectional Encoder Representations from Transformers) natural language processing model to vectorize question Q1 and the summary. The information processing device 100 may also calculate a Euclidean distance based on the first and second vectors, and calculate a score indicating the degree of relevance to the question based on the calculated Euclidean distance. The information processing device 100 may also determine that index information with a calculated score above a predetermined threshold is related to user U1's question. The information processing device 100 may extract index information that it has determined to be related to the question from user U1.

[0028] The information processing device 100 generates an answer to user U1's question Q1 from the extracted index information. Specifically, the information processing device 100 generates a prompt for generating an answer to question Q1 from the extracted index information and inputs it into a text generation AI such as a GPT model to generate an answer to question Q1 (step S204). To give a specific example, based on the extracted index information, the information processing device 100 may generate a prompt like the following from question Q1, which is something like "What is the maximum amount for accommodation expenses in the company's accommodation regulations?" and a summary such as "Business trip expenses are capped at 5,000 yen per night."

[0029] #instruction: You are a professional editor. Please answer the question based on the following contextual information. Please write your answer in the language of the question. #Context information: Business trip expenses are capped at 5,000 yen per night. #question: What is the maximum accommodation expense limit according to our company's accommodation policy?

[0030] By generating a prompt like the one described above and inputting it into the text generation AI, an answer such as "According to the business trip expense regulations, the maximum accommodation cost per night is 5,000 yen" is output. The information processing device 100 then provides the answer to question Q1 to user U1 (step S205). Specifically, the information processing device 100 sends information regarding the answer to question Q1 (information for displaying the answer) to the terminal device 10. The terminal device 10 then displays the answer provided by the information processing device 100.

[0031] As described above, in Embodiment 2, when the information processing device 100 receives a question Q1 from user U1, it narrows down the index information by category and label, extracts relevant index information, and generates a prompt for the text generation AI using a summary of the extracted index information. If the information used to generate the answer (index information) is not narrowed down, the answer may be generated based on information unrelated to question Q1, and therefore may not be able to generate an appropriate answer to the user's question. Also, if the information used to generate the answer (index information) is not narrowed down, it is necessary to refer to a lot of index information when generating the answer, so it takes a long time to generate the answer. For this reason, according to the information processing device 100 of this embodiment, by narrowing down the information used to generate the answer (index information) based on the category and label associated with question Q1, an appropriate answer to the user's question can be provided in a short time.

[0032] User U1, for example, enters question Q1 into a bulletin board of a specific tool. For example, User U1 enters question Q1 after specifying a predetermined category and label. When the information processing device 100 receives question Q1 from user U1, it generates a prompt using a text generation AI, but in doing so, it generates the prompt using a library in which the content is divided into regions and index information is generated for each region. In this case, for example, the information processing device 100 calculates a score indicating the degree of relevance for each index information and extracts index information with a high relevance score. The information processing device 100 may calculate the Euclidean distance based on the first vector and the second vector, and calculate the relevance score based on the calculated Euclidean distance. Then, for example, the information processing device 100 generates a prompt to generate an answer using the text data of the summary of the extracted index information. Alternatively, if there is no index information with a high relevance score, the information processing device 100 generates a prompt to generate an answer without using index information. In this case, for example, an answer to question Q1 is generated without index information (summary, etc.). In this case, for example, the information processing device 100 generates an answer to question Q1 by inputting a prompt to the text generation AI to generate an answer without using index information. The information processing device 100 may, for example, terminate information processing without generating a prompt to generate an answer if there is no index information with a high relevance score.

[0033] (content) Figure 5 shows an example of content managed in a company according to this embodiment. The company has multiple departments, and multiple content items are created for each department and managed on the company's servers. In the example shown in Figure 5, "○○ Corporation" has multiple departments such as the "Legal Department," "Human Resources Department," and "General Affairs Department." The "Legal Department" also creates multiple content items, such as travel expense regulations and employment regulations. In this embodiment, each department within the company (Legal Department, Human Resources Department, General Affairs Department, etc.) corresponds to a "category," the titles of the content created by each department (regulations, rules, manuals, etc.) correspond to "labels," and each area within the content (Article 1, Article 2, Article 3, etc.) corresponds to a scope label.

[0034] (Chatbot screen) Figure 6 shows an example of the home screen 200 of a chatbot according to an embodiment. When a user launches the chatbot application using the terminal device 10, the home screen 200 shown in Figure 6 is displayed on the terminal device 10. The chatbot's home screen 200 displays a description of the chatbot and a filter button B1. The filter button B1 is a button that displays a window for entering a question to the chatbot, as well as a category and a label.

[0035] Figure 7 shows an example of a chatbot window 210 according to the embodiment. Window 210 is displayed on the terminal device 10 when the user selects (taps) the filter button B1. Window 210 displays a category input area 211, a label input area 212, a message input area 213, a cancel button B2, and a send button B3.

[0036] The category input area 211 is an area for the user to enter the category of the question. For example, when the user selects (tap) the category input area 211, a dropdown menu of category options (e.g., Legal Department, Human Resources Department, General Affairs Department) may be displayed. When the user selects one of several categories, the selected category may be displayed in the category input area 211. Alternatively, the user may directly enter the category into the category input area 211 using a text input application.

[0037] The label input area 212 is an area for the user to enter a label for a question. For example, when the user selects (tap) the label input area 212, a dropdown menu of label options corresponding to the selected category (e.g., travel expense regulations, employment regulations) may be displayed. When the user selects one of several labels, the selected label may be displayed in the label input area 212. Alternatively, the user may directly enter a label into the label input area 212 using a text input application.

[0038] The message input area 213 is an area for the user to enter the content of a question. For example, the user may select (tap) the message input area 213 and enter the content of the question using a text input application. For example, the question could be "What is the maximum amount for accommodation expenses in the company's accommodation regulations?"

[0039] The Cancel button B2 is used to cancel input to window 210. When the user selects (tap) the Cancel button B2, window 210 will be closed. The Send button B3 is used to instruct the user to send their question. When the user selects (tap) the Send button B3, the terminal device 10 associates the category entered in the category input area 211, the label entered in the label input area 212, and the content of the question entered in the message input area 213, and sends this information to the information processing device 100.

[0040] Furthermore, window 210 may be provided with a range label input area for entering range labels (such as Article 1, Article 2, Article 3, etc.). In this case, when the user selects (tap) the send button B3, the terminal device 10 may associate the category entered in the category input area 211, the label entered in the label input area 212, the range label entered in the range label input area, and the content of the question entered in the message input area 213, and send this information to the information processing device 100.

[0041] Figure 8 shows an example of an internal inquiry form 220 according to the embodiment. Users may enter their inquiry content not only through the window 210 shown in Figure 7, but also through the internal inquiry form 220 shown in Figure 8. When a user accesses the information processing device 100 using the terminal device 10, the internal inquiry form 220 shown in Figure 8 is displayed on the terminal device 10. The internal inquiry form 220 displays a question input area 221, a company selection area 222, a department name input area 223, a branch name input area 224, a category input area 225, a label input area 226, a contact information input area 227, a cancel button B4, and a send button B5.

[0042] The question input area 221 is an area for the user to enter the content of a question. For example, the user may select (tap) the question input area 221 and enter the content of the question using a text input application. For example, the question could be "What is the maximum amount for accommodation expenses in the company's accommodation regulations?"

[0043] The company name input area 222 is for entering the name of the company to which the user belongs. In the example shown in Figure 8, the options "○○ Corporation", "△△ Corporation", and "□□ Corporation" are displayed. These companies may be group companies or affiliated companies. Also, in the example shown in Figure 8, "○○ Corporation" is shown to be selected.

[0044] The department name input area 223 is an area for entering the name of the department to which the user belongs. For example, the user may select (tap) the department name input area 223 and enter the department name using a text input application.

[0045] The location name input area 224 is an area for entering the name of the location to which the user belongs. For example, when the user selects (tap) the location name input area 224, a dropdown menu may be displayed with multiple location name options (e.g., XX Office, YY Office, ZZ Office, ZZ Office). When the user selects one of the multiple location names, the selected location name may be displayed in the location name input area 224. Alternatively, the user may directly enter the location name into the location name input area 224 using a text input application.

[0046] The category input area 225 is an area for the user to enter the category of the question. For example, when the user selects (tap) the category input area 225, a dropdown menu of category options (e.g., Legal Department, Human Resources Department, General Affairs Department) may be displayed. When the user selects one of several categories, the selected category may be displayed in the category input area 225. Alternatively, the user may directly enter the category into the category input area 225 using a text input application.

[0047] The label input area 226 is an area for the user to enter a label for a question. For example, when the user selects (tap) the label input area 226, a dropdown menu of label options corresponding to the selected category (e.g., travel expense regulations, employment regulations) may be displayed. When the user selects one of several labels, the selected label may be displayed in the label input area 226. Alternatively, the user may directly enter a label into the label input area 226 using a text input application.

[0048] The contact input area 227 is an area for entering the user's email address as a contact. For example, the user may select (tap) the contact input area 227 and enter their contact information (email address) using a text input application. The chatbot's response may be sent to the email address entered in the contact input area 227.

[0049] The Cancel button B4 is used to cancel input into the internal inquiry form 220. When the user selects (tap) the Cancel button B4, the internal inquiry form 220 is closed. The Send button B5 is used to instruct the user to send their question. When the user selects (tap) the Send button B5, the terminal device 10 associates the question text (inquiry content) entered in the question text input area 221, the affiliated company entered in the affiliated company selection area 222, the department name entered in the department name input area 223, the location name entered in the location name input area 224, the category entered in the category input area 225, and the label entered in the label input area 226, and sends this information to the information processing device 100.

[0050] Furthermore, the internal inquiry form 220 may include a range label input area for entering range labels (e.g., Article 1, Article 2, Article 3). In this case, when the user selects (taps) the send button B5, the terminal device 10 may associate the question text (inquiry content) entered in the question text input area 221, the affiliated company entered in the affiliated company selection area 222, the department name entered in the department name input area 223, the location name entered in the location name input area 224, the category entered in the category input area 225, the label entered in the label input area 226, and the range label entered in the range label input area, and send this information to the information processing device 100.

[0051] Furthermore, in user inquiries, inputting categories and labels is optional, and the category input area 225 and label input area 226 may be omitted from the internal inquiry form 220. For example, an internal inquiry form 220 may be provided for each type of inquiry, and categories and labels may be associated with each internal inquiry form 220. This allows for the automatic determination of categories and labels according to the internal inquiry form 220 used, thereby reducing the burden on users to input information into the internal inquiry form 220.

[0052] Traditionally, within a company, various departments freely created different documents, making it difficult to know where certain information was located. This resulted in inefficient operations due to the need for inquiries about inquiries themselves (for example, inquiries about what to inquire about, where and how to send an inquiry request, or whether an inquiry is even necessary). In contrast, this embodiment indexes internal documents by associating them with categories and labels, and uses the index information to generate answers to user questions using a data generation AI. This allows for the generation of appropriate answers to user questions, thereby improving operational efficiency.

[0053] (Variations in processing) In the above embodiment, even if the information processing device 100 determines that there is index information relevant to the question, the index information accessible to the user may be restricted. For example, the information processing device 100 may extract relevant index information from the index information accessible to the user and generate a prompt.

[0054] In the above embodiment, the information processing device 100 may perform processing to optimize the range (size of the area) of the area division according to the accuracy of the answer. In the example of the above embodiment, if a user evaluates an answer to a question such as "What is the upper limit for accommodation expenses in the company's accommodation regulations?", such as "According to the business trip expense regulations, the upper limit for accommodation expenses per night is 5,000 yen," the information processing device 100 may accept the user's evaluation as an evaluation of that answer. In this way, the information processing device 100 may accept evaluations from (multiple) users, and if the evaluation is low, it may set up multiple index information with changed area division ranges, and if changing to one of the multiple index information would improve the evaluation, it may change to the index information that gives the highest evaluation.

[0055] In the above embodiment, if a model answer (model response) such as a Q&A (or FAQ) is predetermined, the information processing device 100 may generate a prompt using the model answer. For example, the information processing device 100 may generate a prompt that instructs the output of the model answer along with the answer according to the above embodiment. For example, the information processing device 100 may generate a prompt that outputs a model answer such as "The model answer is as follows: ○×○×..." along with an answer such as "According to the travel expense regulations, the maximum amount for accommodation per night is 5,000 yen."

[0056] In the above embodiment, the information processing device 100 may link to the generated response. For example, the information processing device 100 may link to the content corresponding to the label associated with the index information used to generate the response.

[0057] In the above embodiment, examples were given of labels such as "Business Trip Expense Regulations," "Employment Regulations," and "Childcare / Family Care Leave Regulations," but the embodiment is not limited to these examples. The label in the embodiment may be a label indicating a specific regulation or law, a specific rule or manual, but is not limited to these examples. Also, in the above embodiment, an example was given of a scope label such as "Article 5," but the embodiment is not limited to this example. The scope label in the embodiment may be a label indicating a specific chapter, article, or sentence, but is not limited to these examples.

[0058] [4. Configuration of terminal equipment] Next, the configuration of the terminal device 10 according to the embodiment will be described using Figure 9. Figure 9 is a diagram showing an example of the configuration of the terminal device 10 according to the embodiment. As shown in Figure 9, the terminal device 10 has a communication unit 11, an input unit 12, an output unit 13, and a control unit 14.

[0059] (Communications Section 11) The communication unit 11 is implemented, for example, by a NIC (Network Interface Card). The communication unit 11 is connected to a predetermined network N by wire or wireless connection and sends and receives information to and from the information processing device 100 via the predetermined network N.

[0060] (Input section 12) The input unit 12 accepts various operations from the user. In Figure 4 above, it accepts various operations from user U1. For example, the input unit 12 may accept various operations from the user via a display surface using a touch panel function. Alternatively, the input unit 12 may accept various operations from buttons provided on the terminal device 10, or from a keyboard or mouse connected to the terminal device 10. For example, the input unit 12 accepts an operation to input a question.

[0061] (Output section 13) The output unit 13 is a display screen for a tablet terminal, for example, which is implemented using a liquid crystal display or an organic EL (Electro-Luminescence) display, and is a display device for displaying various types of information. For example, the output unit 13 displays information transmitted from the information processing device 100. For example, the output unit 13 displays the answer to a user question generated by the information processing device 100.

[0062] (Control Unit 14) The control unit 14 is, for example, a controller, and is implemented by a CPU (Central Processing Unit) or MPU (Micro Processing Unit) executing various programs stored in the memory device inside the terminal device 10 using RAM (Random Access Memory) as the working area. For example, these various programs include application programs installed on the terminal device 10. For example, these various programs include application programs that display information transmitted from the information processing device 100 (such as answers to user questions). The control unit 14 is also implemented by an integrated circuit such as an ASIC (Application Specific Integrated Circuit) or an FPGA (Field Programmable Gate Array).

[0063] As shown in Figure 9, the control unit 14 has a receiving unit 141 and a transmitting unit 142, and realizes or executes the information processing operations described below.

[0064] (Receiver 141) The receiving unit 141 receives information transmitted from, for example, the information processing device 100. For example, the receiving unit 141 receives information related to the answer to the user's question. For example, the receiving unit 141 receives information for displaying the answer to the user's question.

[0065] (Transmitter 142) The transmitting unit 142 transmits, for example, information about operations performed by the user. For example, the transmitting unit 142 transmits information about questions received from the user (for example, questions entered by the user, specified by the user, or selected by the user).

[0066] [5. Configuration of the Information Processing Device] Next, the configuration of the information processing device 100 according to the embodiment will be described using Figure 10. Figure 10 is a diagram showing an example of the configuration of the information processing device 100 according to the embodiment. As shown in Figure 10, the information processing device 100 has a communication unit 110, a storage unit 120, and a control unit 130. The information processing device 100 may also have an input unit (for example, a keyboard or mouse) that receives various operations from the administrator of the information processing device 100, and a display unit (for example, a liquid crystal display) for displaying various information.

[0067] (Communications Department 110) The communication unit 110 is implemented, for example, by a NIC. The communication unit 110 is connected to the network N by wire or wireless connection and sends and receives information to and from terminal devices 10, etc., via the network N.

[0068] (Storage unit 120) The storage unit 120 is implemented by, for example, semiconductor memory elements such as RAM and flash memory, or storage devices such as hard disks and optical discs. As shown in Figure 10, the storage unit 120 has an index information storage unit 121 and a response evaluation information storage unit 122.

[0069] The index information storage unit 121 stores index information. Here, Figure 11 shows an example of the index information storage unit 121 according to the embodiment. The information stored in the index information storage unit 121 is used, for example, to generate answers. As shown in Figure 11, the index information storage unit 121 has items such as "index information ID", "content ID", "category", "label", "range label", and "summary".

[0070] "Index Information ID" indicates identification information used to identify index information. "Content ID" indicates identification information used to identify content. "Category" indicates the category (a category indicating the type of content). "Label" indicates the label (a label indicating the content of the content). "Scope Label" indicates the scope label (a label indicating a specific scope within the content). "Summary" indicates the summary (a summary corresponding to a specific scope within the content).

[0071] The response evaluation information storage unit 122 stores evaluation information for the responses. Here, Figure 12 shows an example of the response evaluation information storage unit 122 according to the embodiment. The information stored in the response evaluation information storage unit 122 is used, for example, to optimize the range of area division. As shown in Figure 12, the response evaluation information storage unit 122 has items such as "index information ID" and "response evaluation information".

[0072] The "Index Information ID" indicates identification information used to identify the index information. The "Response Evaluation Information" indicates evaluation information for the response. In the example shown in Figure 12, conceptual information such as "Response Evaluation Information #1" and "Response Evaluation Information #2" is stored in "Response Evaluation Information," but in reality, a score indicating the overall evaluation by (multiple) users is stored.

[0073] (Control unit 130) The control unit 130 is a controller, and is implemented, for example, by a CPU or MPU executing various programs stored in the memory device inside the information processing device 100 using RAM as the working area. Alternatively, the control unit 130 can be implemented by an integrated circuit such as an ASIC or FPGA.

[0074] As shown in Figure 10, the control unit 130 includes an acquisition unit 131, a division unit 132, a first generation unit 133, a second generation unit 134, a reception unit 135, a narrowing unit 136, an extraction unit 137, a third generation unit 138, a providing unit 139, and a vectorization unit 140, and realizes or executes the information processing operations described below. Note that the internal configuration of the control unit 130 is not limited to the configuration shown in Figure 10, and other configurations are also acceptable as long as they perform the information processing described later.

[0075] (Acquisition part 131) The acquisition unit 131 acquires various information from the storage unit 120. The acquisition unit 131 also stores the acquired information in the storage unit 120.

[0076] The acquisition unit 131 acquires various information from external information processing devices. The acquisition unit 131 also acquires various information from other information processing devices such as the terminal device 10.

[0077] The acquisition unit 131 acquires, for example, the content that is the subject of a question from the user. For example, the acquisition unit 131 acquires the text data of the content (for example, the entire text).

[0078] (Divided part 132) The division unit 132 divides the content acquired by the acquisition unit 131, for example. For example, the division unit 132 divides the content acquired by the acquisition unit 131 into contextually appropriate areas (for example, areas for each chapter, each article, each sentence, etc.). The division unit 132 divides the content into one or more (or more) areas in this way, for example.

[0079] (First generation unit 133) The first generation unit 133 generates a summary for each region divided by the division unit 132, for example. For example, if the content is divided into regions for each chapter, the first generation unit 133 generates a summary for each chapter.

[0080] (Second generation unit 134) The second generation unit 134 generates index information that links, for example, the summary generated by the first generation unit 133, the category of the content from which the summary is to be compiled, and the label corresponding to the content from which the summary is to be compiled. The second generation unit 134 also generates index information that links, for example, the summary generated by the first generation unit 133, the category of the content from which the summary is to be compiled, the label corresponding to the content from which the summary is to be compiled, and the range label corresponding to the range within the content from which the summary is to be compiled (the range corresponding to the summary generated by the first generation unit 133).

[0081] The index information generated in this way is used to generate prompts for producing answers when a user submits a question. Furthermore, this index information is used to extract index information relevant to a question when a user submits a question.

[0082] (Reception desk 135) The reception unit 135 receives questions from users, for example, that are associated with categories and labels. The reception unit 135 also receives questions from users, for example, that are associated with categories, labels, and range labels.

[0083] (Filtering section 136) The filtering unit 136 narrows down the index information based on, for example, the category and label associated with the question received by the reception unit 135. The extraction unit 137 also narrows down the index information based on, for example, the category, label, and range label associated with the question received by the reception unit 135. For example, if the question received by the reception unit 135 is associated with the category "Legal Department" and the label "Business Trip Expense Regulations", the filtering unit 136 narrows down the index information based on the conditions that the category is "Legal Department" and the label is "Business Trip Expense Regulations". Also, for example, if the question received by the reception unit 135 is associated with the category "Legal Department", the label "Business Trip Expense Regulations", and the range label "Article 5", the filtering unit 136 narrows down the index information based on the conditions that the category is "Legal Department", the label is "Business Trip Expense Regulations", and the range label is "Article 5".

[0084] (Extraction part 137) The extraction unit 137 extracts index information relevant to the question from the index information narrowed down by the filtering unit 136, for example. In this process, the extraction unit 137 calculates a score indicating the degree of relevance to the question for each piece of index information, and extracts index information whose score is above a predetermined threshold. For example, the extraction unit 137 may generate a first vector vectorized from the user's question and a second vector vectorized from the summary linked to the index information, and determine whether or not the index information is relevant to the user's question by comparing the first and second vectors. For example, the extraction unit 137 may use the BERT natural language processing model to vectorize the question and summary. The extraction unit 137 may also calculate a Euclidean distance based on the first and second vectors, calculate a score indicating the degree of relevance to the question based on the calculated Euclidean distance, and extract index information whose score is above a predetermined threshold. Furthermore, the extraction unit 137 does not extract index information if, for example, there is no index information with a score above a predetermined threshold (i.e., there is no highly relevant index information).

[0085] (Third generation unit 138) The third generation unit 138 generates answers to questions received from the user, for example, from the index information extracted by the extraction unit 137. For example, the third generation unit 138 generates a prompt for generating answers to questions received from the user, and the answer output by inputting that prompt into the text generation AI is used as the answer to the question received from the user.

[0086] The third generation unit 138 generates answers to user questions without using index information if, for example, the extraction unit 137 does not extract index information. For example, the third generation unit 138 generates a prompt to generate answers to user questions without using index information, and the response output by inputting that prompt to the text generation AI is used as the answer to the user question.

[0087] (Providing Department 139) The providing unit 139 provides, for example, information regarding the answer generated by the third generation unit 138 to the user who asked the question. For example, the providing unit 139 transmits content showing the answer generated by the third generation unit 138 to the terminal device 10. Also, for example, the providing unit 139 transmits information to the terminal device 10 for displaying the content showing the answer generated by the third generation unit 138.

[0088] [6. Information Processing Flow] Next, the information processing procedure by the information processing system 1 according to the embodiment will be described using Figures 13 and 14. Figures 13 and 14 are flowcharts showing the information processing procedure by the information processing system 1 according to the embodiments (Embodiment 1 and Embodiment 2).

[0089] As shown in Figure 13 (Embodiment 1), the information processing device 100 acquires the content to be learned (step S301).

[0090] The information processing device 100 divides the acquired content into context-appropriate regions (step S302).

[0091] The information processing device 100 generates a summary for each divided region (step S303).

[0092] The information processing device 100 generates index information corresponding to the generated summary (step S304).

[0093] As shown in Figure 14 (Embodiment 2), the information processing device 100 receives a question from the user (step S401).

[0094] The information processing device 100 narrows down the index information based on the category and label associated with the received question (step S402).

[0095] The information processing device 100 determines whether or not there is index information related to the received question (step S403).

[0096] If the information processing device 100 determines that there is index information relevant to the received question (step S403; YES), it extracts index information that is highly relevant to the received question (step S404). Then, the information processing device 100 generates an answer to the question based on the extracted index information (step S405).

[0097] On the other hand, if the information processing device 100 determines that there is no index information related to the received question (step S403; NO), it generates an answer to the question without extracting index information (step S406).

[0098] [7. Effects] As described above, the information processing device 100 according to the embodiment includes an acquisition unit 131, a division unit 132, a first generation unit 133, a second generation unit 134, a reception unit 135, a filtering unit 136, an extraction unit 137, and a third generation unit 138. The acquisition unit 131 acquires content that is the subject of a question from the user. The division unit 132 divides the content acquired by the acquisition unit 131 into a plurality of regions according to the context. The first generation unit 133 generates a summary for each of the plurality of regions divided by the division unit 132. The second generation unit 134 generates index information that links the summary generated by the first generation unit 133 with the category of the content from which the summary is to be summarized and a label indicating the content from which the summary is to be summarized. The reception unit 135 receives questions from the user that are associated with categories and labels. The filtering unit 136 filters the index information based on the categories and labels associated with the questions received by the reception unit 135. The extraction unit 137 extracts index information relevant to the question from the index information narrowed down by the filtering unit 136. The third generation unit 138 generates an answer to the question based on the index information extracted by the extraction unit 137.

[0099] As a result, the information processing device 100 according to this embodiment can appropriately acquire information (index information) used to generate answers, and therefore can provide appropriate answers to user questions in a short amount of time.

[0100] Furthermore, the vectorization unit 140 generates a first vector, which is a vectorized version of the question, and a second vector, which is a vectorized version of the summary. The extraction unit 137 compares the first vector and the second vector to extract index information that is relevant to the question.

[0101] As a result, the information processing device 100 according to this embodiment can, for example, extract index information relevant to a user's question with high accuracy.

[0102] Furthermore, the third generation unit 138 generates prompts for generating answers to questions based on the extracted index information, and outputs the responses by inputting the prompts into the text generation AI (trained model) to generate answers to questions from the user.

[0103] As a result, the information processing device 100 according to the embodiment can, for example, generate an appropriate answer to a question from a user.

[0104] Furthermore, the second generation unit 134 generates index information that links the summary, category, label, and range label indicating the scope within the content from which the summary is derived.

[0105] As a result, the information processing device 100 according to the embodiment can be structured, for example, to make data easier to organize and search.

[0106] Furthermore, the second generation unit 134 generates index information that associates summaries, categories, and labels for each region of a size determined based on the evaluation of the responses.

[0107] As a result, the information processing device 100 according to the embodiment can, for example, optimize the range of region division according to the accuracy of the response.

[0108] Furthermore, the content is the terms of service.

[0109] As a result, the information processing device 100 according to the embodiment can, for example, generate an appropriate response regarding the rules.

[0110] [8. Hardware Configuration] Furthermore, the information processing device 100 according to the above-described embodiment is realized by a computer 1000 having the configuration shown in Figure 15. Figure 15 is a hardware configuration diagram showing an example of a computer that realizes the functions of the information processing device 100. The computer 1000 has a CPU 1100, RAM 1200, ROM 1300, HDD 1400, communication interface (I / F) 1500, input / output interface (I / F) 1600, and media interface (I / F) 1700.

[0111] The CPU 1100 operates based on programs stored in the ROM 1300 or HDD 1400, controlling various components. The ROM 1300 stores boot programs executed by the CPU 1100 when the computer 1000 starts up, as well as programs that depend on the computer 1000's hardware.

[0112] The HDD1400 stores programs executed by the CPU1100, as well as data used by such programs. The communication interface1500 acquires data from other devices via a predetermined communication network and sends it to the CPU1100, and transmits data generated by the CPU1100 to other devices via the predetermined communication network.

[0113] The CPU 1100 controls output devices such as displays and printers, and input devices such as keyboards and mice, via the input / output interface 1600. The CPU 1100 acquires data from input devices via the input / output interface 1600. The CPU 1100 also outputs the generated data to output devices via the input / output interface 1600.

[0114] The media interface 1700 reads a program or data stored in the recording medium 1800 and provides it to the CPU 1100 via the RAM 1200. The CPU 1100 loads the program from the recording medium 1800 onto the RAM 1200 via the media interface 1700 and executes the loaded program. The recording medium 1800 can be, for example, an optical recording medium such as a DVD (Digital Versatile Disc) or PD (Phase Change Rewritable Disk), a magneto-optical recording medium such as an MO (Magneto-Optical disk), tape media, magnetic recording medium, or semiconductor memory.

[0115] For example, when the computer 1000 functions as an information processing device 100 according to the embodiment, the CPU 1100 of the computer 1000 realizes the functions of the control unit 130 by executing a program loaded on the RAM 1200. The CPU 1100 of the computer 1000 reads and executes these programs from the recording medium 1800, but as another example, these programs may be obtained from other devices via a predetermined communication network.

[0116] [9. Other] Furthermore, among the processes described in the above embodiments, all or part of the processes described as being performed automatically can be performed manually, or all or part of the processes described as being performed manually can be performed automatically by known methods. In addition, the processing procedures, specific names, and various data and parameters shown in the above document and drawings can be changed at will unless otherwise specified. For example, the various information shown in each figure is not limited to the information shown.

[0117] Furthermore, the components of each illustrated device are functionally conceptual and do not necessarily need to be physically configured as shown. In other words, the specific forms of distribution and integration of each device are not limited to those shown, and all or part of them can be functionally or physically distributed and integrated in any unit according to various loads and usage conditions.

[0118] Furthermore, the embodiments described above can be combined as appropriate, as long as the processing content is not contradictory.

[0119] Although some embodiments of the present invention have been described in detail above with reference to the drawings, these are illustrative examples, and the present invention can be implemented in various other forms with modifications and improvements based on the knowledge of those skilled in the art, starting with the embodiments described in the disclosure section of the invention.

[0120] Furthermore, the terms "section, module, unit" mentioned above can be replaced with "means" or "circuit." For example, the acquisition unit can be replaced with acquisition means or acquisition circuit. [Explanation of symbols]

[0121] 1. Information Processing System 10 Terminal devices 11 Communications Department 12 Input section 13 Output section 14 Control Unit 100 Information Processing Devices 110 Communications Department 120 Storage section 121 Index Information Storage Unit 122 Answer Evaluation Information Storage Unit 130 Control Unit 131 Acquisition Department 132 Division 133 1st generation part 134 Second generation part 135 Reception Department 136 Restriction section 137 Extraction part 138 Third generation part 139 Provision Department 140 Vectorization section 141 Receiving Unit 142 Transmitter N Network

Claims

1. A unit that retrieves content that is the subject of a user's question, A division unit that divides the aforementioned content into multiple regions according to the context, A first generation unit that generates summaries for each of the aforementioned multiple regions, A second generation unit generates index information that links the aforementioned summary, the category of the content from which the summary is derived, and a label indicating the content from which the summary is derived. A reception unit that receives questions from the user that are associated with the aforementioned categories and labels, A filtering unit that narrows down the index information based on the category and label associated with the question, An extraction unit extracts index information relevant to the question from the narrowed-down index information, A third generation unit generates an answer to the question based on the extracted index information, An information processing device equipped with the following features.

2. The system further includes a vectorization unit that generates a first vector obtained by vectorizing the aforementioned question and a second vector obtained by vectorizing the aforementioned summary. The extraction unit is By comparing the first vector and the second vector, the index information relevant to the question is extracted. The information processing apparatus according to claim 1.

3. The third generation unit is, Based on the extracted index information, a prompt is generated to produce an answer to the question. The response output by inputting the aforementioned prompt into the trained model is used to generate an answer to the user's question. The information processing apparatus according to claim 1.

4. The second generation unit is, The index information is generated by linking the summary, the category, the label, and the range label indicating the scope within the content from which the summary is derived. The information processing apparatus according to claim 1.

5. The second generation unit is, Based on the evaluation of the above response, the index information is generated, which is associated with the summary, category, and label for each region of a size determined by the above response. The information processing apparatus according to claim 1.

6. The aforementioned content is subject to the terms and conditions. The information processing apparatus according to claim 1.

7. Information processing device, Retrieve the content that is the subject of the user's question, The aforementioned content is divided into multiple regions according to the context, A summary is generated for each of the aforementioned multiple areas, An index is generated that links the summary, the category of the content from which the summary is derived, and a label indicating the content from which the summary is derived. Questions associated with the aforementioned categories and labels are received from the user. Based on the category and label associated with the question, the index information is narrowed down. From the narrowed-down index information, extract the index information relevant to the question, Based on the extracted index information, an answer to the question is generated. Information processing methods.

8. In an information processing device, The content that the user is asking about will be retrieved. The aforementioned content is divided into multiple regions according to the context, A summary is generated for each of the aforementioned multiple areas. An index information is generated that links the aforementioned summary, the category of the content from which the summary is derived, and a label indicating the content from which the summary is derived. The user submits questions associated with the aforementioned categories and labels. Based on the category and label associated with the aforementioned question, the index information is narrowed down. From the narrowed-down index information, extract the index information relevant to the question. Based on the extracted index information, the system generates an answer to the question. program.

Images