Information processing device, information processing method, and program

Abstract

Provided are an information processing device, an information processing method, and a program that are capable of constructing and using a question and answer database based on text data that can also be used to improve the display of a paywall screen. [Solution] In an information processing system 400, an information processing device 1C includes an acquisition unit that acquires text data, an input unit that inputs the text data and an instruction to output a question whose answer is contained in the text data to a language model, and a storage unit that stores a pair of a question generated from the language model and the answer as question and answer data in a question and answer database 3.

Description

[Technical field]

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

[0002] When providing articles to users on the Internet, a common technique is to display a paywall screen that encourages them to register as a member or purchase the article (Figure 1). The paywall screen displays part of the beginning of the article, allowing the user to infer information that can be obtained by viewing the remainder of the article, stimulating the user's desire to register or purchase (hereinafter, collectively referred to as "desire to purchase"). If the user wishes to read the rest of the article, they can purchase the article, or in Figure 1, register as a free member, which allows them to view the full text of the article. Summary of the Invention [Problem to be solved by the invention]

[0003] However, if the number of characters displayed at the beginning of the article is increased on the paywall screen to allow the user to concretely imagine the content of the article, the remaining part of the article is reduced, and the information obtained by viewing the article is reduced, which reduces the user's willingness to purchase. On the other hand, if the number of characters displayed at the beginning of the article is reduced on the paywall screen, it becomes difficult for the user to imagine the content of the article and to infer the information obtained by viewing the article, which causes a problem of reducing the user's willingness to purchase. For this reason, a method for improving the user's willingness to purchase on a paywall screen is required.

[0004] The inventor came up with the idea that, instead of displaying a part of the article on a paywall screen, questions can be created for information that can be obtained by viewing the article content, and the questions can be displayed on the paywall screen, thereby allowing the user to imagine the information that can be obtained by viewing the article, and the problem of the length of displaying a part of the article can be resolved. The inventor also came up with the idea that by creating a database of pairs of questions and answers created based on the article and referring to this, in addition to improving the display on the paywall screen, it can be utilized for displaying Q&A related to the article, completing queries in searches, displaying snippets when answering searches, etc.

[0005] Therefore, an object of the present disclosure is to provide an information processing device capable of constructing a question and answer database based on text data that can also be used to improve the display of a paywall screen, and an information processing device that utilizes the question and answer database. [Means for solving the problem]

[0006] In order to achieve the above object, the information processing device of the present disclosure includes a text data acquisition unit that acquires text data, an input unit that inputs the text data and an instruction to output a question whose answer is present in the text data into a language model; The system further comprises a storage unit for storing a set of the question generated from the language model and the answer in a question and answer database as question and answer data.

[0007] The information processing device of the present disclosure includes a receiving unit that receives identification information of document data; a text data acquisition unit that acquires text data associated with the identification information of the text data from a text database; a question and answer data acquisition unit that acquires question and answer data linked to identification information of the text data from a question and answer database; a transmission unit for transmitting the text data and the question and answer data, the question and answer database includes a plurality of question and answer data; The question and answer data is data that links a question obtained by inputting model sentence data and an instruction to output a question whose answer exists in the model sentence data into a language model with the answer in the sentence.

[0008] The information processing device of the present disclosure includes a receiving unit that receives identification information of document data; a text data acquisition unit that acquires text data associated with the identification information of the text data from a text database; a question and answer data acquisition unit that acquires question and answer data related to the text data from a question and answer database; a transmission unit for transmitting the text data and the question and answer data, the question and answer database includes a plurality of question and answer data; The question and answer data is data that links a question obtained by inputting model sentence data and an instruction to output a question whose answer exists in the model sentence data into a language model with the answer in the sentence.

[0009] The information processing device of the present disclosure includes a search query receiving unit that receives a search query; a question and answer data acquisition unit that acquires question and answer data related to the search query from a question and answer database; a transmission unit for transmitting the question and answer data; the question and answer database includes a plurality of question and answer data; The question and answer data is data that links a question obtained by inputting model sentence data and an instruction to output a question whose answer exists in the model sentence data into a language model with the answer in the sentence.

[0010] The information processing device of the present disclosure includes a search query receiving unit that receives a search query; a question and answer data acquisition unit that acquires question and answer data related to the search query from a question and answer database; a search result acquisition unit that acquires search results from a search server using the search query; a transmission unit for transmitting the search results and the question and answer data; the question and answer database includes a plurality of question and answer data; The question and answer data is data that links a question obtained by inputting model sentence data and an instruction to output a question whose answer exists in the model sentence data into a language model with the answer in the sentence. Effect of the Invention

[0011] According to the present disclosure, it is possible to provide an information processing device capable of constructing a question and answer database based on text data that can also be used to improve the display of a paywall screen, and an information processing device that uses the question and answer database. [Brief description of the drawings]

[0012] [Figure 1] FIG. 1 is a diagram showing an example of a paywall screen that is displayed before purchasing an article. [Diagram 2] FIG. 2 is a block diagram showing an example of a processing system including the information processing device according to the first embodiment. [Diagram 3] FIG. 3 is a block diagram illustrating an example of a hardware configuration of the information processing apparatus according to the first embodiment. [Figure 4] FIG. 4 is a flowchart showing an example of the information processing method and program according to the first embodiment. [Diagram 5] FIG. 5 is a block diagram showing an example of a processing system including an information processing device according to the second embodiment. [Figure 6] FIG. 6 is a flowchart showing an example of an information processing method and program according to the second embodiment. [Figure 7] FIG. 7 is a block diagram showing an example of a processing system including an information processing device according to the third embodiment. [Figure 8] FIG. 8 is a flowchart showing an example of an information processing method and program according to the third embodiment. [Figure 9]FIG. 9 is a block diagram showing an example of a processing system including an information processing device according to the fourth embodiment. [Figure 10] FIG. 10 is a flowchart showing an example of an information processing method and program according to the fourth embodiment. [Figure 11] FIG. 11 is a block diagram showing an example of a processing system including an information processing device according to the fifth embodiment. [Figure 12] FIG. 12 is a flowchart showing an example of an information processing method and program according to the fifth embodiment. [Figure 13] FIG. 13 is a diagram showing an example of a paywall screen displayed on the display unit of the user terminal according to the fifth embodiment. [Figure 14] FIG. 14 is a block diagram showing an example of a processing system including an information processing device according to the sixth embodiment. [Figure 15] FIG. 15 is a flowchart showing an example of an information processing method and program according to the sixth embodiment. [Figure 16] FIG. 16 is a block diagram showing an example of a processing system including an information processing device according to the seventh embodiment. [Figure 17] FIG. 17 is a flowchart showing an example of an information processing method and program according to the seventh embodiment. [Figure 18] FIG. 18 is a block diagram showing an example of a processing system including an information processing device according to the eighth embodiment. [Figure 19] FIG. 19 is a flowchart showing an example of an information processing method and program according to the eighth embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0013] Hereinafter, the present disclosure will be described in detail with reference to the drawings, giving examples. However, the present disclosure is not limited by the following description. In addition, in the following Figs. 2 to 19, the same parts may be given the same reference numerals, and the description thereof may be omitted. In addition, in the drawings, for convenience of explanation, the structure of each part may be appropriately simplified and the dimensional ratio of each part may be shown diagrammatically, different from the actual one. In addition, each embodiment and description can be mutually referred to unless otherwise specified. In addition, when the expression "~" is used in this specification, it is used to mean including the numerical value or physical value before and after it. In addition, in this specification, the expression "A and / or B" includes "A only", "B only", and "both A and B".

[0014] (Embodiment 1) The present embodiment is an example of an information processing system including an information processing device of the present disclosure. As shown in FIG. 2, the information processing system 100 of the present embodiment includes an information processing device 1, a language model 2 outside the system, and a question and answer database (Q&A DB) 3. The information processing device 1 includes an acquisition unit 11, an input unit 12, and a storage unit 13. As shown in FIG. 2, in the information processing system 100, the information processing device 1 can be connected (communicated) with the language model 2 and the Q&A DB 3 in one direction or two directions via a communication line network 4. In the information processing system 100 of the first embodiment, the information processing device 1 is directly connected to the communication line network 4 outside the system, but may be connected to the communication line network 4 via an authentication server (not shown) outside the system. In addition, the number of the language model 2 and the Q&A DB 3 that can be connected to the information processing device 1 is one each, but may be more than one each. In the present embodiment, the information processing device 1 is configured as a server in which the program of the present disclosure is installed.

[0015] The information processing device 1 may be incorporated in a server or the like as a personal computer (PC) in which the program of the present disclosure is installed or as a system. The information processing device 1 of the present embodiment may be a system composed of one or more computers or servers capable of executing the program of the present disclosure, that is, a cloud computing system. The personal computer may constitute a computer cluster. Although not shown, the information processing device 1 may be configured to be connectable to an external terminal of a system administrator via the authentication server and the communication line network 4, so that the system administrator can manage the information processing device 1 from the external terminal. In the present embodiment, the information processing system 100 includes one information processing device 1, but may include multiple information processing devices 1.

[0016] The language model 2 may be, for example, a server in which a large-scale language model (LLM) such as Generative Pre-trained Transformer (GPT (registered trademark)) (Open AI) such as GPT-4o, GPT-4, GPT-3.5, etc.; PaLM2, LaMDA2, Google Bard, Gemini (Google Inc.); Claude-3 and other Glaude (Anthropic); OpenCALM (CyberAgents); Alpaca (Stanford University) such as Alpaca 7B; Dolly-v2 (Databricks); RWKV, Raven (BlinkDL); StableLM, StableLM-Tuned (Stability.ai); or a large-scale multimodal model (LMM), Phi, etc., is stored. In the information processing device 1 of the first embodiment, the language model 2 is a server in which a large-scale language model is stored.

[0017] The Q&A DB 3 is a database server in which pairs of questions and answers corresponding to each piece of text data 106 are linked and stored.

[0018] The communication network 4 is not particularly limited and may be a known network, for example, a wired or wireless network. Examples of the communication network 4 include the Internet, WWW (World Wide Web), telephone lines, LAN (Local Area Network), and WiFi (Wireless Fidelity).

[0019] 3 illustrates a block diagram of a hardware configuration of the information processing device 1. The information processing device 1 includes, for example, a CPU (Central Processing Unit) 101, which is a computing element, a memory 102, a bus 103, a storage device 104, an input device 107, a display 108, and a communication device (communication unit) 109. Each unit of the information processing device 1 is connected via the bus 103 by its respective interface (I / F). The hardware configurations of the language model 2 and the Q&A DB 3 are also similar to that of the information processing device 1, and the description thereof is hereby incorporated by reference.

[0020] The CPU 101 cooperates with other components through a controller (system controller, I / O controller, etc.) and controls the entire information processing device 1. In the information processing device 1, the CPU 101 executes the program 105 of the present disclosure, text data 106, and other programs, and also reads or writes various information. Specifically, in this embodiment, the CPU 101 functions as an acquisition unit 11, an input unit 12, and a storage unit 13. The information processing device 1 includes the CPU 101 as a calculation device (calculation element), but may include other calculation devices such as a GPU (Graphics Processing Unit) and an APU (Accelerated Processing Unit), or may include a combination of the CPU and these. Note that the CPU 101 functions as each unit other than the storage unit in other embodiments, for example.

[0021] The memory 102 includes a main memory. The main memory is also called a primary storage device. When the CPU 101 performs processing, the memory 102 reads various operation programs, such as a program 105 of the present disclosure, stored in a storage device 104 (auxiliary storage device) described below. The CPU 101 then reads data from the memory 102, decodes it, and executes the program. The main memory is a RAM (random access memory). The memory 102 may further include a ROM (read only memory).

[0022] The bus 103 can also be connected to an external device. Examples of the external device include an external storage device (external database, etc.) and a printer. The information processing device 1 can be connected to a communication network 4 by a communication device 109 connected to the bus 103, and can also be connected to external devices such as a language model 2 and a Q&A DB 3 via the communication network 4.

[0023] The storage device 104 is also called an auxiliary storage device, for example, in contrast to the main memory (primary storage device). As described above, the storage device 104 includes an operating program including the program 105 of the present disclosure. The storage device 104 may store data including text data 106 and the like. The storage device 104 includes, for example, a storage medium and a drive for reading and writing data from and to the storage medium. The storage medium is not particularly limited, and may be, for example, an internal or external type, and may be, for example, a hard disk (HD), a floppy disk (FD), a CD-ROM, a CD-R, a CD-RW, an MO, a DVD, a flash memory, a memory card, and the like, and the drive is not particularly limited. The storage device 104 may be, for example, a hard disk drive (HDD) in which the storage medium and the drive are integrated.

[0024] The information processing device 1 further includes an input device 107 and a display 108 which is an output device. Examples of the input device 107 include pointing devices such as a touch panel, a track pad, and a mouse; a keyboard; an imaging means such as a camera and a scanner; a card reader such as an IC card reader and a magnetic card reader; and a voice input means such as a microphone. Examples of the display 108 include a display device such as an LED (light emitting diode) display and a liquid crystal display. In this embodiment, the input device 107 and the display 108 are configured separately, but the input device 107 and the display 108 may be configured as an integrated device, such as a touch panel display. In addition, in the information processing device 1, the input device 107 and the display 108 are of any configuration, and may be omitted or may include one or more of them.

[0025] Next, an example of processing in the information processing system 100 of this embodiment will be described using the flowchart of Fig. 4, taking as an example a case where the text data is a newspaper article. As shown in Fig. 4, the information processing method and program of the first embodiment performs the processes of S1 (acquisition), S2 (generation), and S3 (storage) in this order.

[0026] In step S1, the acquisition unit 11 acquires text data included in a newspaper article (S1). Specifically, the acquisition unit 11 acquires text data 106 of a newspaper article (including a title) stored in the storage device 104 of the information processing device 1 as model text data to be used for generating question and answer data. The number of pieces of text data 106 (model text data) acquired by the acquisition unit 11 is plural, for example, 1 to 1000000000 pieces, 100 to 1000000000 pieces, or 1000 to 500000000 pieces. By increasing the number of pieces of text data 106 acquired by the acquisition unit 11, the number of pieces of question and answer data stored in the Q&A DB3 can be increased. The text data 106 in the Q&A DB3 is data in which identification information (ID, identifier) ​​of a newspaper article is linked to the text of the newspaper article. The text data 106 in the Q&A DB3 may further be linked to tag information of the type of newspaper article. In this embodiment, the text data 106 is stored in the storage device 104 of the information processing device 1, but the text data 106 may be stored in a text database (article database) connectable to the information processing device 1 via an authentication server outside the system and the communication line network 4. In this case, the acquisition unit 11 acquires the newspaper article text data 106 from the text database (for example, a text DB3A described below). The acquisition unit 11 may acquire all of the stored newspaper article text data 106, or may acquire the newspaper article text data 106 by excluding a portion of it in consideration of the number of characters and a social viewpoint.

[0027] Next, in step S2, the input unit 12 inputs the text data 106 and an instruction (prompt) to output a question whose answer exists in the text data 106 to the language model 2. Specifically, in step S2, the input unit 12 first links the text data 106 (model text data) acquired by the acquisition unit 11 to a prompt (instruction) for outputting a pair of a question whose answer exists in the text data 106 and the answer. Then, in step S2, the input unit 12 inputs the text data 106 (model text data) and the prompt linked to the text data 106 as a pair to the language model 2 (S2). As a result, the input unit 12 generates a pair of a question whose answer exists in a newspaper article and the answer by using the large-scale language model stored in the language model 2. As a specific example, when the article in FIG. 1 is used as text data 106 and is input together with the prompt to the language model 2, for example, as shown in FIG. 13(B), a pair of the question and answer is output, which is a question "Q1. What is Legal Tech?" and an answer "A1. Legal Tech is a service that supports the legal department of a company in the creation and review of contracts and case management. Startups have appeared since the mid-2010s, and the style of providing services to customers via the cloud has become widespread." The question and the answer may be one-to-one, many-to-one, one-to-many, or many-to-many. The S2 step may be performed multiple times for the same text data 106 by changing the prompt. In this case, the pair of the question and the answer whose answer exists in the newspaper article generated in the S2 step may hold the results of multiple generation. As a result, in the S2 step, for example, a pair in which one answer is linked to multiple questions, a pair in which multiple answers are linked to one question, and a pair in which multiple answers are linked to multiple questions can be generated.

[0028] Next, in step S3, the storage unit 13 links the question generated by the language model 2 in step S2 with the answer in the text data 106 to form a pair, and stores them as question and answer data in the Q&A DB 3. The storage unit 13 also links the ID of the newspaper article used to generate the question and answer in the above storage to the question-and-answer pair (question and answer data). The information processing device 1 then generates question-and-answer pairs and stores them in the Q&A DB 3 in the same way for all of the text data 106 acquired by the acquisition unit 11, and terminates the process.

[0029] In the information processing device 1 of the first embodiment, the text data 106 is used to generate questions whose answers exist in the text data 106 by using a large-scale language model stored in the language model 2, and the questions and answers are linked and stored as question and answer data in the Q&A DB 3. Therefore, according to the information processing device 1 of the first embodiment, questions and answers that are in line with the contents of the text data 106 can be generated, and therefore, for example, a Q&A DB 3 can be constructed in which pairs of questions and answers suitable for displaying the paywall screen of the text data 106 are stored.

[0030] In the information processing device 1 of the first embodiment, the generation of the question and answer database based on the text data 106 of newspaper articles has been described as an example, but the present disclosure is not limited thereto, and the question and answer database may be generated based on text data other than newspaper articles. Examples of the text data other than newspaper articles include magazine articles, books, official documents, web media, reports, financial statements, securities reports, and press releases. The text data 106 other than newspapers may be, for example, videos such as news; images such as charts accompanying news or reports; text in which the contents of the videos or images are transcribed (converted) into text media; and the like. When the video is used as the text data other than newspapers, the language model 2 may be a language model that supports the input of videos, for example, Gemini Pro (Google Inc.), etc.

[0031] The information processing device 1 of the first embodiment may include, for example, an indexing unit that assigns indexes to the text data 106 and the question and answer pair (question and answer data). In this case, the index in the storage unit 13 is, for example, a search index used during a search. The index can be assigned by storing the index in a database having a search function, such as a normal full-text search engine or a vector database. The assignment of the index can also be referred to as index linking.

[0032] (Embodiment 2) This embodiment is another example of the information processing device of the present disclosure. FIG. 5 is a block diagram showing an information processing system 200 including an information processing device 1A of this embodiment. As shown in FIG. 5, the information processing device 1A of this embodiment includes an extraction unit 14 in addition to the configuration of the information processing device 1 of the first embodiment. The hardware configuration of the information processing device 1A is the same as that of the information processing device 1 of FIG. 3, except that the CPU 101 includes the configuration of the information processing device 1A of FIG. 5 instead of the configuration of the information processing device 1 of FIG. 3. Except for these points, the configuration of the information processing device 1A of the second embodiment is the same as the configuration of the information processing device 1 of the first embodiment, and the description thereof can be used.

[0033] Next, an example of processing of the information processing system 200 including the information processing device 1A of the present embodiment will be described with reference to the flowchart of Fig. 6. As shown in Fig. 6, the information processing method of the second embodiment includes step S4 (extraction) in addition to steps S1 to S3 of the information processing method of the first embodiment.

[0034] First, the information processing system 200 of this embodiment performs step S1 in the same manner as the information processing system 100 of the first embodiment.

[0035] Next, in step S2, the input unit 12 inputs the text data 106 (model text data) acquired by the acquisition unit 11 and a prompt for outputting a question whose answer exists in the text data 106 to the language model 2 (S2). Here, in the information processing device 1 of the first embodiment, the input unit 12 links the text data 106 with a prompt for outputting a pair of a question and an answer so that the answer exists in the text data 106, and inputs the link to the language model 2. In contrast, in the information processing device 1A of the second embodiment, the input unit 12 links the text data 106 with a prompt for outputting a question whose answer exists in the text data 106, and inputs the link to the language model 2. For this reason, in the information processing device 1 of the first embodiment, a pair of a question and an answer for the text data 106 is output from the language model 2, whereas in the information processing device 1A of the second embodiment, only a question for the text data 106 is output from the language model 2. Therefore, after step S2, the information processing apparatus 1A of the second embodiment extracts an answer to the question obtained in step S2 from the text data 106 in step S4 described later.

[0036] In step S4, the extraction unit 14 extracts an answer to the question generated by the language model 2 in step S2 from the text data 106. Specifically, the extraction unit 14 can extract the answer using a semantic search, using the text data 106 used to generate the question and the question.

[0037] Then, in step S3, the storage unit 13 links the question generated by the language model in step S2 with the answer extracted in step S4, and stores them as question and answer data in the Q&A DB 3. Then, the information processing device 1A similarly generates question-answer pairs and stores them in the Q&A DB 3 for all of the text data 106 acquired by the acquisition unit 11, and terminates the process.

[0038] The information processing device 1A of the second embodiment outputs the question from the question-answer pair by using the language model 2, and extracts the answer from the text data 106. Therefore, the information processing device 1A of the second embodiment can reduce calculation costs compared to the case where the question-answer pair is output from the language model 2.

[0039] (Embodiment 3) This embodiment is another example of the information processing device of the present disclosure. FIG. 7 is a block diagram showing an information processing system 300 including an information processing device 1B of this embodiment. As shown in FIG. 7, the information processing device 1B of this embodiment includes a type determination unit 15 in addition to the configuration of the information processing device 1 of the first embodiment. The hardware configuration of the information processing device 1B is the same as that of the information processing device 1 of FIG. 3, except that the CPU 101 includes the configuration of the information processing device 1B of FIG. 7 instead of the configuration of the information processing device 1 of FIG. 3. Except for these points, the configuration of the information processing device 1B of the third embodiment is the same as the configuration of the information processing device 1 of the first embodiment, and the description thereof can be used.

[0040] Next, an example of processing of the information processing system 300 including the information processing device 1B of the present embodiment will be described with reference to the flowchart of Fig. 8. As shown in Fig. 8, the information processing method of the third embodiment includes step S5 (type determination) in addition to steps S1 to S3 of the information processing method of the first embodiment.

[0041] First, the information processing system 300 of this embodiment performs step S1 in the same manner as the information processing system 100 of the first embodiment.

[0042] Next, in step S5, the type discrimination unit 15 discriminates the type of the text data 106 (S5). The type of the text data 106 is, for example, a classification based on the theme of the description of the text data 106. If the text data 106 is a newspaper article, the type of the text data 106 is the genre (article classification) of the newspaper article. The genre of the newspaper article can be, for example, politics, economy, culture, science, society, disaster, crime, international, sports, etc. The type of the text data 106 may be, for example, a type given by the creator of the text data 106, or a type based on a classification term in which a unique keyword is assigned to the newspaper article (for example, a classification term of Nikkei Telecom, etc.). If the type data is linked to the text data 106, the type discrimination unit 15 discriminates the type of the text data 106 based on the type data linked to the text data 106. The type discrimination unit 15 may discriminate the type of the text data 106 based on, for example, language processing. In this case, the language processing may be, for example, a model using a neural network such as Naive Bayes, Support Vector Machine (SVM), Random Forest, or BERT (Bidirectional Encoder Representations from Transformers (Google Inc.)).

[0043] Next, in step S2, the input unit 12 inputs the sentence data 106 (model sentence data) acquired by the acquisition unit 11 and a prompt set according to the type of sentence data 106 for outputting a pair of a question to which an answer exists in the sentence data 106 and the answer, to the language model 2 (S2). Specifically, according to the type of sentence data 106, appropriate positive examples and / or inappropriate negative examples are prepared as the question, and linked to the type of sentence data 106. Then, in step S2, the input unit 12 links the sentence data 106 with the prompt, as well as positive examples and / or negative examples corresponding to the type of sentence data 106 determined in step S5, and inputs the sentence data 106 to the language model 2.

[0044] Next, the information processing system 300 of this embodiment performs step S3 in the same manner as the information processing system 100 of embodiment 1. Then, the information processing device 1B similarly generates question-answer pairs and stores them in the Q&A DB3 for all of the text data 106 acquired by the acquisition unit 11, and ends the process.

[0045] In the information processing device 1B of the third embodiment, the input unit 12 inputs the question in association with a positive example and / or a negative example when inputting the question to the language model 2. Therefore, the information processing device 1B can generate question-answer pairs that match the content or type of the text data 106, and can therefore build a Q&A DB3 that stores question and answer pairs that are more suitable for displaying the paywall screen of the text data 106, for example.

[0046] In the information processing device 1B of the third embodiment, the positive examples and / or negative examples of the questions are linked to the type of text data 106, but the present disclosure is not limited to this, and positive examples and / or negative examples of questions regardless of the type of text data 106 may be prepared, linked to the text data 106 and the prompt, and input to the language model 2. This enables the information processing device of the present disclosure to improve the quality of the questions.

[0047] (Embodiment 4) This embodiment is another example of the information processing device of the present disclosure. FIG. 9 is a block diagram showing an information processing system 400 including an information processing device 1C of this embodiment. As shown in FIG. 9, the information processing device 1C of this embodiment includes a number setting unit 16 and a question determination unit 17 in addition to the configuration of the information processing device 1 of the first embodiment. The hardware configuration of the information processing device 1C is the same as that of the information processing device 1 of FIG. 3, except that the CPU 101 includes the configuration of the information processing device 1C of FIG. 9 instead of the configuration of the information processing device 1 of FIG. 3. Except for these points, the configuration of the information processing device 1C of the fourth embodiment is the same as the configuration of the information processing device 1 of the first embodiment, and the description thereof can be used.

[0048] Next, an example of processing of the information processing system 400 including the information processing device 1C of the present embodiment will be described with reference to the flowchart of Fig. 10. As shown in Fig. 10, the information processing method of the fourth embodiment includes a step S6 (number setting) and a step S7 (question determination) in addition to the steps S1 to S3 of the information processing method of the first embodiment.

[0049] First, the information processing system 400 of this embodiment performs step S1 in the same manner as the information processing system 100 of the first embodiment.

[0050] Next, in step S6, the number setting unit 16 sets the number of questions to be output, that is, the number of question-answer pairs, based on the number of characters of the text data (S6). The number of question and answer pairs to be generated by the language model 2 for one piece of text data 106 can be adjusted, for example, according to the number of characters of the input text data 106. Specifically, the number of pairs can be set to be relatively large when the number of characters of the text data 106 is relatively large, and the number of question and answer pairs to be generated can be set to be relatively small when the number of characters of the text data 106 is relatively small. As a specific example, when the number of characters of the text data 106 is 10 to 1000 characters, the number of pairs can be set to be, for example, 1 to 15 pairs, when the number of characters is 40 to 300 characters, the number of pairs can be set to be, for example, 1 to 5 pairs, or when the number of characters is 50 to 90 characters, the number of pairs can be set to be, for example, 1 to 2 pairs.

[0051] Next, in step S2, the input unit 12 inputs to the language model 2 the sentence data 106 (model sentence data) acquired by the acquisition unit 11, a prompt to output pairs of questions and answers whose answers exist in the sentence data 106, and the number of questions to be output (the number of question-answer pairs) (S2).

[0052] Next, in step S7, the question determination unit 17 determines whether the question generated from the language model 2 satisfies a pre-specified evaluation criterion (S7). The evaluation criterion may be, for example, the number of characters in the question, or the accuracy of the subject, predicate, and modifier. The accuracy may be determined, for example, by syntactically analyzing the presence of 5W1H in the question. The syntactic analysis may be performed, for example, by KNP (http: / / nlp.ist.i.kyoto-u.ac.jp / ?KNP). The evaluation criterion may be, for example, whether the question is a question likely to be used by a user of the Q&A DB3. Then, in step S7, if the question generated from the language model 2 does not satisfy the pre-specified evaluation criterion, that is, if the result is No, and if there is no unprocessed text data 106, the information processing device 1C ends the process. On the other hand, if the question generated from the language model 2 satisfies the pre-specified evaluation criterion, i.e., if Yes, the process proceeds to step S3, where the storage unit 13 stores the question and answer pairs determined in step S7 to satisfy the evaluation criterion as question and answer data in the Q&A DB 3. The information processing device 1C then similarly generates question-answer pairs and stores them in the Q&A DB 3 for all of the text data 106 acquired by the acquisition unit 11, and terminates the process.

[0053] In the information processing device 1C of the fourth embodiment, the number of question-answer pairs to be output is determined according to the number of characters in the text data 106. When the number of characters in the text data 106 is fixed, the quality of the questions in the question-answer pairs output from the language model 2 relatively decreases when the number of outputs exceeds a certain level. On the other hand, in the information processing device 1C of the fourth embodiment, the number setting unit 16 can determine the number of question-answer pairs to be output from the language model 2 according to the number of characters in the text data 106, so that the deterioration of the quality of the questions in the question-answer pairs output from the language model 2 can be suppressed. In addition, in the information processing device 1C of the fourth embodiment, it is determined whether the questions output from the language model 2 satisfy an evaluation criterion, and the question-answer pairs that satisfy the evaluation criterion are stored in the Q&A DB3. Therefore, according to the information processing device 1C of the fourth embodiment, the quality of the questions stored in the Q&A DB3 can be improved. Therefore, the information processing device 1C of the fourth embodiment can build a Q&A DB3 in which pairs of questions and answers that are more suitable for displaying the paywall screen of the text data 106 are stored, for example.

[0054] The information processing devices of the first to fourth embodiments may be combined with the information processing devices of the fifth to eighth embodiments described below to configure a system. In this case, each device may be configured as a single device, or may be configured to be connectable (communicable) with each other unidirectionally or bidirectionally via a communication line network.

[0055] (Embodiment 5) This embodiment is an example of an information processing system of the present disclosure capable of displaying questions in a Q&A DB on a paywall screen. FIG. 11 is a block diagram showing an information processing system 600 including an information processing device 6 of this embodiment. As shown in FIG. 11, the information processing system 600 includes a Q&A DB 3, a text database (text DB) 3A, and an information processing device 6. The information processing device 6 includes a receiving unit 61, a text data acquisition unit 62, a question and answer data acquisition unit 63, a transmitting unit 64, a purchase instruction determination unit 65, and a text data transmitting unit 66. The Q&A DB 3, the text database (text DB) 3A, and the information processing device 6 are connectable (communicable) with each other in one direction or both directions via a communication line network 4 outside the system. As shown in FIG. 11, the user terminal 5 is connectable (communicable) with the information processing device 6 in one direction or both directions via the communication line network 4. Note that, although there are two user terminals 5 in FIG. 11, there may be one or more. In this embodiment, the information processing device 6 is configured as a server in which the program of the present disclosure is installed.

[0056] The Q&A DB3 can be constructed in the same manner as in the embodiments 1 to 4. In this embodiment, the Q&A DB3 is the Q&A DB constructed in the embodiment 1. The Q&A DB3 stores a plurality of question-answer pairs created from text data 106 in a text DB3A described later.

[0057] The text DB 3A is a database in which a plurality of text data 106 are stored.

[0058] Examples of the user terminal 5 include a personal computer (PC); a mobile terminal such as a mobile phone, a smartphone, or a tablet terminal; a smart watch, a smart glass, or a wearable terminal; and the like.

[0059] The information processing device 6 may be incorporated in a server or the like as a personal computer (PC) in which the program of the present disclosure is installed or as a system. The information processing device 6 may also be a system consisting of one or more computers or servers capable of executing the program of the present disclosure, that is, a cloud computing system. The personal computer may constitute a computer cluster. Although not shown, the information processing device 6 may be configured to be connectable to an external terminal of a system administrator via the authentication server and the communication line network 4, so that the system administrator can manage the information processing device 6 from the external terminal. In this embodiment, the information processing system 600 includes one information processing device 6, but may include multiple information processing devices.

[0060] The hardware configuration of the information processing device 6 is the same as that of the information processing device 1 in FIG. 3, except that the CPU 101 has the configuration of the information processing device 6 in FIG. 11 instead of the configuration of the information processing device 1 in FIG.

[0061] Next, an example of processing in the information processing system 600 of this embodiment will be described with reference to the flowchart of Fig. 12, taking the case where the text data is a newspaper article as an example. As shown in Fig. 12, the information processing method and program of the fifth embodiment performs each of the processes of S8 (selection of text data), S9 (transmission of identification information), S10 (reception of identification information), S11 (acquisition of text data), S12 (acquisition of question and answer data), S13 (transmission of data), S14 (display of data), S15 (purchase decision), S16 (transmission of instruction data), S17 (reception decision), and S18 (transmission of text data) in this order.

[0062] First, in step S8, the user operates the user terminal 5 to select a newspaper article (text data 106) to be viewed. Next, in step S9, the communication unit of the user terminal 5 transmits to the information processing device 6 the identification information of the newspaper article linked to the selected newspaper article (text data 106).

[0063] In step S10, the receiving unit 61 receives the identification information of the text data 106. Next, in step S11, the text data acquiring unit 62 acquires the text data 106 linked to the identification information of the text data 106 from the text DB 3A. Specifically, the text data acquiring unit 62 inputs the identification information of the text data 106 from which information is to be acquired to the text DB 3A via the communication line network 4. The text DB 3A extracts the corresponding text data 106 from the identification information and transmits it to the information processing device 6. Then, the text data acquiring unit 62 acquires the transmitted text data 106.

[0064] Next, in step S12, the question and answer data acquiring unit 63 acquires question and answer data linked to the identification information of the text data 106 from the Q&A DB 3. Specifically, the question and answer data acquiring unit 63 inputs the identification information of the text data 106 from which information is to be acquired to the Q&A DB 3 via the communication network 4. The Q&A DB 3 extracts the question and answer data linked to the corresponding identification information from the identification information, and transmits it to the information processing device 6. Then, the question and answer data acquiring unit 63 acquires the transmitted question and answer data.

[0065] Next, in step S13, the transmission unit 64 transmits the text data 106 obtained in step S11, the question and answer data obtained in step S12, and purchase necessity data for confirming whether or not to purchase the text data 106, to the user terminal 5. Here, the transmission unit 64 may transmit a part of the beginning of the text data 106. The text data 106 to be transmitted is, for example, the first paragraph at the beginning, specifically, 10 to 150 characters. Furthermore, the transmission unit 64 transmits data of the question portion to be displayed on the user terminal 5, out of the question and answer data. The number of questions to be transmitted is, for example, 1 to 10, or 4 to 8. The transmission unit 64 may transmit display instruction data for instructing how to display the part of the text data 106 and the question and answer data.

[0066] The communication unit of the user terminal 5 receives the question and answer data and the purchase necessity data. When a part of the text data 106 is transmitted from the transmission unit 64, the communication unit of the user terminal 5 further receives the part of the text data 106 transmitted from the transmission unit 64. Then, as shown in FIG. 13(A), the user terminal 5 displays the questions (Q1 to Q5 in the figure) of the question and answer data and a part of the text data 106 (the article title in the figure, "Paper contracts that will never disappear: Two trends in legal tech") on the display unit (display) of the user terminal 5 (S14). When the transmission unit 64 transmits the display instruction data, the communication unit of the user terminal 5 receives the display instruction data, and the user terminal 5 displays the questions of the question and answer data and the text data 106 on the display unit (display) of the user terminal 5 based on the display instruction data. The user terminal 5 also displays a message ("Register as a free member" in the figure) to confirm whether or not the text data 106 needs to be purchased, and determines whether or not there is a purchase instruction from the user (S15). In this embodiment, the purchase instruction is an instruction to enter registration information for a free member and to confirm the registration. If the user purchases the text data 106, that is, if the answer is Yes, the user terminal 5 transmits purchase instruction data instructing the purchase of the text data 106 to the information processing device 6 via the communication network 4 (S16). On the other hand, if the user does not purchase the text data 106, that is, if the answer is No, the user terminal 5 proceeds to step S17 without performing any further processing.

[0067] Next, in step S17, the purchase instruction determination unit 65 determines whether or not there is a purchase instruction for the text data 106. If the information processing device 6 has not received the purchase instruction data from the user terminal 5, that is, if the answer is No, the information processing device 6 ends the process. On the other hand, if the information processing device 6 has received the purchase instruction data from the user terminal 5, that is, if the answer is Yes, the remaining or all of the text data 106 is transmitted to the user terminal 5 (S18). On the user terminal 5, as shown in FIG. 13(B), the questions (Q1 to Q5 in the figure) of the question and answer data and the entire text of the text data 106 are displayed. In the display, as shown in FIG. 13(B), the answer portion may be highlighted or the like so that the answer portion to the question in the question and answer data can be grasped. In addition, in the display, as shown in FIGS. 13(A) and (B), the display may be displayed in a form in which a link for moving to the answer portion corresponding to the question is added to the display portion of the question. Then, the information processing device 6 ends the process.

[0068] In the information processing device 6 of the fifth embodiment, the question and answer data acquisition unit 63 acquires a question-answer pair corresponding to the newspaper article (text data 106) from the Q&A DB 3 based on the newspaper article that the user wishes to view. Therefore, in the information processing device 6 of the fifth embodiment, in addition to a part of the newspaper article (text data 106), a question of a question-answer pair created from the newspaper article can be displayed on the user terminal 5. Therefore, the user of the user terminal 5 can guess from the question what kind of information he or she may be able to obtain by viewing the remainder of the newspaper article, the part of which is being displayed, thereby increasing the user's willingness to purchase the newspaper article.

[0069] In the information processing device 6 of embodiment 5, a membership-type paywall in which membership registration is required to view articles has been described as an example, but the information processing device of the present disclosure is not limited to this and can also be used for other paywalls, such as a freemium-type paywall in which a paid purchase is required to view additional articles or content, a pay-per-use paywall in which a paid purchase is required depending on the number of views, and a timer-type paywall in which articles can be viewed for a certain period of time but a paid purchase is required to view articles after the certain period has elapsed.

[0070] (Embodiment 6) This embodiment is an example of an information processing system of the present disclosure capable of displaying related questions in a Q&A DB when displaying text data. FIG. 14 is a block diagram showing an information processing system 700 including an information processing device 7 of this embodiment. As shown in FIG. 14, the information processing system 700 includes a Q&A DB3, a text database (text DB) 3A, and an information processing device 7. The information processing device 7 includes a receiving unit 71, a text data acquisition unit 72, a question and answer data acquisition unit 73, and a transmitting unit 74. The Q&A DB3, the text database (text DB) 3A, and the information processing device 7 are connectable (communicable) with each other in one direction or two directions via a communication line network 4 outside the system. As shown in FIG. 14, the user terminal 5 is connectable (communicable) with the information processing device 7 in one direction or two directions via the communication line network 4. In this embodiment, the information processing device 7 is configured as a server in which the program of the present disclosure is installed. In the information processing system 700 of the sixth embodiment, the Q&A DB 3, the text DB 3A, and the user terminal 5 have the same configurations as those of the information processing system 600 of the fifth embodiment, and the description thereof can be cited.

[0071] The hardware configuration of the information processing device 7 is the same as that of the information processing device 1 in FIG. 3, except that the CPU 101 has the configuration of the information processing device 7 in FIG. 14 instead of the configuration of the information processing device 1 in FIG.

[0072] Next, an example of processing in the information processing system 700 of this embodiment will be described using the flowchart of Fig. 15, taking the case where the text data is a newspaper article as an example. As shown in Fig. 15, the information processing method and program of embodiment 6 performs each of the processes of S19 (selection of text data), S20 (transmission of identification information), S21 (reception of identification information), S22 (acquisition of text data), S23 (acquisition of question and answer data), S24 (transmission of data), and S25 (display of data) in this order.

[0073] First, in step S19, the user operates the user terminal 5 to select a newspaper article (text data 106) to be viewed. Next, in step S20, the communication unit of the user terminal 5 transmits, to the information processing device 7, identification information of the newspaper article linked to the selected newspaper article (text data 106).

[0074] In step S21, the receiving unit 71 receives identification information of the text data 106. Next, in step S22, the text data acquiring unit 72 acquires the text data 106 linked to the identification information of the text data 106 from the text DB 3A. Specifically, the text data acquiring unit 72 inputs the identification information of the text data 106 from which information is to be acquired to the text DB 3A via the communication line network 4. The text DB 3A extracts the corresponding text data 106 from the identification information and transmits it to the information processing device 7. Then, the text data acquiring unit 72 acquires the transmitted text data 106.

[0075] Next, in step S23, the question and answer data acquiring unit 73 acquires question and answer data related (similar) to the text data 106 from the Q&A DB3. Specifically, the question and answer data acquiring unit 73 inputs the text data 106 acquired in step S22 in order to acquire question and answer data from the Q&A DB3 via the communication network 4. The Q&A DB3 compares the input text data 106 with the question and answer data stored in the Q&A DB3, and extracts question and answer data related to the text data 106. The extraction can be performed, for example, by converting the input text data 106 and the question and answer data stored in the Q&A DB3 into vectors (document vectors), evaluating the similarity between the vectors, and extracting question and answer data whose similarity is equal to or higher than a certain level. If indexes are assigned to the text data 106 and the question and answer data (if an index is linked to the question and answer data), the Q&A DB3 may extract question and answer data related to the text data 106 by index search, using the index of the input text data 106 and the index of each question and answer data in the Q&A DB3. The Q&A DB3 transmits the extracted question and answer data to the information processing device 7 via the communication network 4. Then, the question and answer data acquisition unit 73 acquires the transmitted question and answer data.

[0076] Next, in step S24, the sending unit 74 sends the text data 106 obtained in step S22 and the question and answer data obtained in step S23 to the user terminal 5. The number of pieces of question and answer data to be sent is, for example, 1 to 10, or 4 to 8.

[0077] The communication unit of the user terminal 5 receives the text data 106 and the question and answer data transmitted from the transmission unit 74. The user terminal 5 displays the text data 106 and the question of the question and answer data on the display unit (display) of the user terminal 5 (S25). The display can be displayed at any position in the text data 106 (newspaper article). As a specific example, the display can be displayed at any position, such as the beginning, end, or middle of the text data 106. Furthermore, when the user selects a question displayed by the user, the user terminal 5 may display the answer of the question and answer data including the question, or may display another newspaper article used to generate the question and answer data. Specifically, when the user has authority to access another newspaper article (e.g., in the case of a paid member), the answer of the question and answer data including the question and the other newspaper article used to generate the question and answer data are displayed. When the user does not have authority to access another newspaper article (e.g., in the case of a member requiring pay-per-use), the other newspaper article used to generate the question and answer data is displayed.

[0078] In the information processing system 700 of the sixth embodiment, question and answer data related to a newspaper article (text data 106) that the user wants to view is acquired, and the question and answer data is displayed on the user terminal 5 together with the newspaper article (text data 106). Therefore, the information processing system 700 of the sixth embodiment can easily provide information related to the user by using the question and answer data. Furthermore, the information processing system 700 of the sixth embodiment can, for example, acquire information derived from a newspaper article (text data 106) related to the newspaper article that the user has viewed by selecting a question after viewing the newspaper article. Therefore, the information processing system 700 of the sixth embodiment can easily acquire information related to information that the user wants to obtain from the newspaper article.

[0079] In the information processing system 700 of the sixth embodiment, the question and answer data acquiring unit 73 compares the text data 106 with the question and answer data stored in the Q&A DB 3, but the present disclosure is not limited to this, and the question and answer data linked to the identification information of the text data 106 may be compared with other question and answer data stored in the Q&A DB 3 to extract related question and answer data. In this way, the information processing system of the present disclosure can more accurately provide the user with information related to the information that the user wants to obtain from a newspaper article.

[0080] (Embodiment 7) The present embodiment is an example of an information processing system of the present disclosure that can extract questions related to a question sentence input in a search query from a Q&A DB and display them as suggestions for a final question sentence in a natural language search. FIG. 16 is a block diagram showing an information processing system 800 including an information processing device 8 of the present embodiment. As shown in FIG. 16, the information processing system 800 includes a Q&A DB 3 and an information processing device 8. The information processing device 8 includes a query receiving unit 81, a question and answer data acquiring unit 82, and a transmitting unit 83. The Q&A DB 3 and the information processing device 8 can be connected (communicated) to each other in one direction or two directions via a communication line network 4 outside the system. As shown in FIG. 16, the user terminal 5 can be connected (communicated) to the information processing device 8 in one direction or two directions via the communication line network 4. In the present embodiment, the information processing device 8 is configured as a server in which the program of the present disclosure is installed. In the information processing system 800 of the seventh embodiment, the Q&A DB 3 and the user terminal 5 have the same configuration as the information processing system 600 of the fifth embodiment, and the description thereof can be used.

[0081] The hardware configuration of information processing device 8 is the same as that of information processing device 1 in FIG. 3, except that CPU 101 has the configuration of information processing device 8 in FIG. 16 instead of the configuration of information processing device 1 in FIG.

[0082] Next, an example of processing in the information processing system 800 of this embodiment will be described with reference to the flowchart in Fig. 17, taking as an example a case where the text data used to construct the Q&A DB3 are newspaper articles. As shown in Fig. 17, the information processing method and program of embodiment 7 performs the processes of S26 (input of search query), S27 (transmission of search query), S28 (reception of search query), S29 (acquisition of question and answer data), S30 (transmission of question and answer data), and S31 (display of question and answer data) in this order.

[0083] First, in step S26, the user operates the user terminal 5 to input a sentence to be searched for (search query) into the search box on the search screen. Next, in step S27, the communication unit of the user terminal 5 transmits the search query to the information processing device 8. The communication unit of the user terminal 5 may transmit the search query while the user is inputting the search query, or may transmit the search query after the user has input the search query, but the former is preferable.

[0084] In step S28, the query receiving unit 81 receives the search query. Next, in step S29, the question and answer data acquiring unit 82 acquires question and answer data related to the search query from the question and answer database. Specifically, the question and answer data acquiring unit 82 inputs the search query acquired from the user terminal 5 to the Q&A DB3 via the communication line network 4 in order to acquire question and answer data. The Q&A DB3 compares the input search query with the question and answer data stored in the Q&A DB3, and extracts question and answer data related to the input search query. The extraction can be performed, for example, by converting the input search query and the question and answer data stored in the Q&A DB3 into vectors (document vectors), evaluating the similarity between the vectors, and extracting question and answer data having a certain level of similarity or higher. If an index is assigned to the question and answer data (if an index is linked to the question and answer data), the Q&A DB 3 may extract question and answer data related to the search query by index search, using the input search query and an index linked to each question and answer data in the Q&A DB 3. The Q&A DB 3 transmits the extracted question and answer data to the information processing device 8 via the communication network 4. Then, the question and answer data acquisition unit 82 acquires the transmitted question and answer data.

[0085] Next, in step S30, the transmission unit 83 transmits the question and answer data obtained in step S29 to the user terminal 5. The number of pieces of question and answer data to be transmitted is, for example, 1 to 10, or 4 to 8.

[0086] The communication unit of the user terminal 5 receives the question and answer data transmitted from the transmission unit 83. The user terminal 5 displays the question and answer data as search query candidates in the search window of the search screen on the display unit (display) of the user terminal 5 (S31). When the user selects a displayed search query candidate, the user terminal 5 may display an answer to the question and answer data including the question. Furthermore, the user terminal 5 may be configured to request the information processing device 8 to transmit text data 106 used to generate the question and answer data used to display the search query candidate, when the user selects a displayed search query candidate.

[0087] In the information processing system 800 of the seventh embodiment, the question and answer data acquisition unit 82 acquires question and answer data related to the search query based on the search query input by the user, and displays the data as search query candidates in the search box of the user terminal 5. This allows the user to easily perform searches by using the search query candidates. In addition, the question and answer data used to create the search query candidates is created from newspaper articles, and answers to the questions of the search query candidates exist in the newspaper articles. In the information processing system 800 of the seventh embodiment, search query candidates for which answers exist can be presented to the user.

[0088] (Embodiment 8) The present embodiment is an example of an information processing system of the present disclosure that can extract question and answer data from a Q&A DB based on a search query input by a user for an emphasized snippet displayed on a search screen, and display the extracted data as an emphasized snippet. FIG. 18 is a block diagram showing an information processing system 900 including an information processing device 9 of the present embodiment. As shown in FIG. 18, the information processing system 900 includes a Q&A DB 3, a search server 3B, and an information processing device 9. The information processing device 9 includes a query receiving unit 91, a question and answer data acquiring unit 92, a search result acquiring unit 93, and a transmitting unit 94. The Q&A DB 3, the search server 3B, and the information processing device 9 can be connected (communicated) with each other in one direction or two directions via a communication line network 4 outside the system. As shown in FIG. 18, the user terminal 5 can be connected (communicated) with the information processing device 9 in one direction or two directions via the communication line network 4. In the present embodiment, the information processing device 9 is configured as a server in which the program of the present disclosure is installed. In the information processing system 900 of the eighth embodiment, the Q&A DB 3 and the user terminal 5 have the same configuration as the information processing system 600 of the fifth embodiment, and the description thereof can be cited.

[0089] The search server 3B is a server having a search engine function, and may be, for example, a commercially available search engine such as Google, Being, Yahoo!, YANDEX, or Baidu.

[0090] The hardware configuration of the information processing device 9 is the same as that of the information processing device 1 in FIG. 3, except that the CPU 101 has the configuration of the information processing device 9 in FIG. 18 instead of the configuration of the information processing device 1 in FIG.

[0091] Next, an example of processing in the information processing system 900 of this embodiment will be described with reference to the flowchart of Fig. 19. As shown in Fig. 19, the information processing method and program of the eighth embodiment performs each of the processes of S32 (input of search query), S33 (transmission of search query), S34 (reception of search query), S35 (acquisition of question and answer data), S36 (acquisition of search results), S37 (transmission of data), and S38 (display of data) in this order.

[0092] First, in step S32, the user operates the user terminal 5 to input a sentence or keyword (search query) to be searched for in the search box on the search screen. Next, in step S33, the communication unit of the user terminal 5 transmits the search query to the information processing device 9. The communication unit of the user terminal 5 may transmit the search query while the user is inputting the search query, or may transmit the search query after the user has input the search query, but the latter is preferable.

[0093] In step S34, the query receiving unit 91 receives the search query. Next, in step S35, the question and answer data acquiring unit 92 acquires question and answer data related to the search query from the question and answer database. Specifically, the question and answer data acquiring unit 92 inputs the search query acquired from the user terminal 5 to the Q&A DB3 via the communication line network 4 in order to acquire question and answer data. The Q&A DB3 compares the input search query with the question and answer data stored in the Q&A DB3, and extracts question and answer data related to the input search query. The extraction can be performed, for example, by converting the input search query and the question and answer data stored in the Q&A DB3 into vectors (document vectors), evaluating the similarity between the vectors, and extracting question and answer data having a certain level of similarity or higher. If an index is assigned to the question and answer data, the Q&A DB3 may extract question and answer data related to the search query by index search using the input search query and an index for each question and answer data in the Q&A DB3. The Q&A DB 3 transmits the extracted question and answer data to the information processing device 9 via the communication network 4. Then, the question and answer data acquisition unit 92 acquires the transmitted question and answer data.

[0094] In step S36, the search result acquisition unit 93 acquires search results from the search server 3B using the search query. The number of search results acquired is, for example, 1 to 10, or 4 to 8.

[0095] Next, in step S37, the transmission unit 94 transmits the question and answer data obtained in step S35 and the search results obtained in step S36 to the user terminal 5. The number of pieces of question and answer data to be transmitted is, for example, 1 to 10, or 4 to 8.

[0096] The communication unit of the user terminal 5 receives the question and answer data and the search result transmitted from the transmission unit 94. The user terminal 5 displays the search result and the question and answer data on a display unit (display) of the user terminal 5 (S38). Furthermore, in step S38, the user terminal 5 displays the question and answer data on the display as a highlighted snippet display.

[0097] In the information processing system 900 of the eighth embodiment, the question and answer data acquisition unit 92 acquires question and answer data related to the search query based on the search query input by the user, and displays the acquired data as a highlighted snippet in the search result display on the user terminal 5. The question and answer data used to display the highlighted snippet is created from a newspaper article. Therefore, the information processing system 900 of the eighth embodiment can present a highly reliable answer to the user's search.

[0098] (Embodiment 9) The program of this embodiment is a program for making a computer execute the information processing method of the present disclosure. In the program of this embodiment, the "processing" can be, for example, a "procedure" or an "instruction". The program of this embodiment may be recorded, for example, in a computer-readable recording medium. The recording medium is, for example, a non-transitory computer-readable storage medium. The recording medium is not particularly limited, and examples thereof include a random access memory (RAM), a read-only memory (ROM), a hard disk (HD), a solid state drive (SSD), an optical disk, and a floppy disk (FD).

[0099] Although the present disclosure has been described above with reference to the embodiments and examples, the present disclosure is not limited to the above-mentioned embodiments and examples. Various modifications that can be understood by a person skilled in the art can be made to the configuration and details of the present disclosure within the scope of the present disclosure.

[0100] <Additional Notes> Some or all of the above-described embodiments and examples may be described as follows, but are not limited to the following: <Paywall screen display> (Appendix 1) A receiving unit for receiving identification information of document data; a text data acquisition unit that acquires text data associated with the identification information of the text data from a text database; a question and answer data acquisition unit that acquires question and answer data linked to identification information of the text data from a question and answer database; a transmission unit for transmitting the text data and the question and answer data, the question and answer database includes a plurality of question and answer data; The question and answer data is data linking a question obtained by inputting model sentence data and an instruction to output a question whose answer exists in the model sentence data into a language model with the answer in the sentence, in an information processing device. (Appendix 2) The transmission unit transmits a part of the text data and purchase necessity data of the text data, a purchase instruction determination unit that determines whether or not a purchase instruction is provided for the text data; a text data transmission unit that transmits the remainder or all of the text data when it is determined that the purchase instruction has been received. 2. An information processing device according to claim 1. (Appendix 3) 3. The information processing device according to claim 1, wherein the transmission unit includes display instruction data for instructing a display of a part of the text data and the question and answer data. <Natural language query> (Appendix 4) a search query receiving unit for receiving a search query; a question and answer data acquisition unit that acquires question and answer data related to the search query from a question and answer database; a transmission unit for transmitting the question and answer data; the question and answer database includes a plurality of question and answer data; The question and answer data is data linking a question obtained by inputting model sentence data and an instruction to output a question whose answer exists in the model sentence data into a language model with the answer in the sentence, in an information processing device. (Appendix 5) The question and answer data is linked to an index, 5. The information processing device according to claim 4, wherein the question and answer data acquisition unit performs an index search using the search query and an index linked to the question and answer data, and acquires question and answer data related to the search query from the question and answer database. <Q&Aデータベース> (Appendix 6) A text data acquisition unit that acquires text data; an input unit that inputs the text data and an instruction to output a question whose answer is present in the text data into a language model; the information processing device comprising: a storage unit configured to store a set of the question generated from the language model and the answer in a question and answer database as question and answer data. (Appendix 7) a type determination unit for determining a type of the text data; The information processing device according to claim 6, wherein the input unit inputs the text data and an instruction to output a question whose answer is present in the text data, which is set according to a type of the text data, to the language model. (Appendix 8) a number setting unit that sets the number of questions to be output by the language model based on the number of characters of the text data; the input unit inputs the text data, an instruction to output a set of questions and answers whose answers are present in the text data, and the number of questions to be output to a language model; The system further includes a question determination unit that determines whether a question generated from the language model satisfies a pre-specified evaluation criterion, 8. The information processing device according to claim 6, wherein the storage unit stores a pair of a question and an answer determined to satisfy the evaluation criterion in the question and answer database as the question and answer data. (Appendix 9) The input unit inputs the text data, an instruction to output a question in which an answer exists in the text data, and a positive example and / or a negative example of the question into the language model, the information processing apparatus according to any one of Appendices 6 to 8. (Appendix 10) The information processing apparatus according to any one of Appendices 6 to 9, comprising an assigning unit that assigns an index to the pair of the question and the answer. (Appendix 11) The input unit inputs the text data and an instruction to output a pair of a question and an answer in which an answer exists in the text data into the language model, The storage unit stores the pair of the question and the answer generated from the language model as the question-and-answer data in the question-and-answer database, the information processing apparatus according to any one of Appendices 6 to 10. (Appendix 12) The information processing apparatus according to any one of Appendices 6 to 11, comprising an extraction unit that extracts an answer to the question generated from the language model from the text data, The storage unit associates the question generated from the language model with the extracted answer and stores them as the question-and-answer data in the question-and-answer database. <Information Processing System> (Appendix 13) An information processing system including the information processing apparatus according to any one of Appendices 1 to 3 and the information processing apparatus according to any one of Appendices 6 to 12. (Appendix 14) An information processing system including the information processing apparatus according to Appendix 4 or 5 and the information processing apparatus according to any one of Appendices 6 to 12. <Q&A Display> (Appendix 15) A receiving unit that receives identification information of text data, A text data acquisition unit that acquires the text data associated with the identification information of the text data from a text database, A question-and-answer data acquisition unit that acquires question-and-answer data related to the text data from a question-and-answer database, a transmission unit for transmitting the text data and the question and answer data, the question and answer database includes a plurality of question and answer data; The question and answer data is data linking a question obtained by inputting model sentence data and an instruction to output a question whose answer exists in the model sentence data into a language model with the answer in the sentence, in an information processing device. <Snippet display> (Appendix 16) a search query receiving unit for receiving a search query; a question and answer data acquisition unit that acquires question and answer data related to the search query from a question and answer database; a search result acquisition unit that acquires search results from a search server using the search query; a transmission unit for transmitting the search results and the question and answer data; the question and answer database includes a plurality of question and answer data; The question and answer data is data linking a question obtained by inputting model sentence data and an instruction to output a question whose answer exists in the model sentence data into a language model with the answer in the sentence, in an information processing device. (Appendix 17) 17. The information processing device according to any one of appendices 1 to 12 and 15 to 16, wherein the language model is a large-scale language model. (Appendix 18) An information processing system comprising the information processing device according to claim 15 or 16 and the information processing device according to any one of claims 6 to 12. <Paywall screen display information processing method> (Appendix 19) a receiving step of receiving identification information of document data; a text data acquisition step of acquiring text data linked to the identification information of the text data from a text database; a question and answer data acquisition step of acquiring question and answer data linked to the identification information of the text data from a question and answer database; a transmitting step of transmitting the text data and the question and answer data; the question and answer database includes a plurality of question and answer data; An information processing method executed on a computer, wherein the question and answer data is data linking a question obtained by inputting model sentence data and an instruction to output a question whose answer exists in the model sentence data into a language model with the answer in the sentence. (Appendix 20) The transmission step includes transmitting a part of the text data and purchase necessity data of the text data, a purchase instruction determination step of determining whether or not the text data includes a purchase instruction; The method described in Appendix 19, further comprising a text data transmission step of transmitting the remainder or all of the text data if it is determined that the purchase instruction exists. (Appendix 21) 21. The method according to claim 19, wherein the transmitting step transmits a part of the text data and display instruction data for instructing the display of the question and answer data. <Method for processing natural language queries> (Appendix 22) a search query receiving step of receiving a search query; a question and answer data acquisition step of acquiring question and answer data related to the search query from a question and answer database; a transmitting step of transmitting the question and answer data; the question and answer database includes a plurality of question and answer data; An information processing method executed on a computer, wherein the question and answer data is data linking a question obtained by inputting model sentence data and an instruction to output a question whose answer exists in the model sentence data into a language model with the answer in the sentence. (Appendix 23) The question and answer data is linked to an index, The question-and-answer data acquisition step is an information processing method described in Supplementary Note 22, which performs index search using the search query and the index associated with the question-and-answer data, and acquires question-and-answer data related to the search query from the question-and-answer database. <Information Processing Method for Q&A Database> (Supplementary Note 24) A sentence data acquisition step of acquiring a sentence, An input step of inputting the sentence data and an instruction to output a question in which an answer exists in the sentence data to a language model, An information processing method executed on a computer, including a storage step of storing, as question-and-answer data, a pair of the question generated from the language model and the answer in a question-and-answer database. (Supplementary Note 25) Including a type determination step of determining the type of the sentence data, The input step is the method described in Supplementary Note 24, which inputs the sentence data and an instruction to output a question in which an answer exists in the sentence data set according to the type of the sentence data to the language model. (Supplementary Note 26) Including a number setting step of setting the number of questions output by the language model based on the number of characters of the sentence data, The input step inputs the sentence data, an instruction to output a set of questions and answers in which an answer exists in the sentence data, and the number of questions to be output to the language model, Furthermore, it includes a question determination step of determining whether the question generated from the language model satisfies a preset evaluation criterion, The storage step is the method described in Supplementary Note 24 or 25, which stores the set of questions and answers determined to satisfy the evaluation criterion as the question-and-answer data in the question-and-answer database. (Supplementary Note 27) The input step is the method described in any one of Supplementary Notes 24 to 26, which inputs the sentence data, an instruction to output a question in which an answer exists in the sentence data, and the positive example and / or negative example of the question to the language model. (Supplementary Note 28) A method according to any one of Appendices 24 to 27, including an indexing step of assigning an index to the pair of the question and the answer. (Appendix 29) The input step inputs the text data and an instruction to output a pair of a question and an answer in which the answer exists in the text data into a language model. The storage step stores the pair of the question and the answer generated from the language model as the question-and-answer data in the question-and-answer database, a method according to any one of Appendices 24 to 28. (Appendix 30) The method includes an extraction step of extracting an answer to the question generated from the language model from the text data. The storage step associates the question generated from the language model with the extracted answer and stores them as the question-and-answer data in the question-and-answer database, a method according to any one of Appendices 24 to 29. (Appendix 31) An information processing method including the information processing method according to any one of Appendices 19 to 21 and the information processing method according to any one of Appendices 24 to 30. (Appendix 32) An information processing method including the information processing method according to Appendix 22 or 23 and the information processing method according to any one of Appendices 24 to 30. <Information Processing Method for Q&A Display> (Appendix 33) A receiving step of receiving identification information of text data. A text data acquisition step of acquiring text data associated with the identification information of the text data from a text database. A question-and-answer data acquisition step of acquiring question-and-answer data related to the text data from a question-and-answer database. A transmission step of transmitting the text data and the question-and-answer data. The question-and-answer database includes a plurality of question-and-answer data. An information processing method executed on a computer, wherein the question and answer data is data linking a question obtained by inputting model sentence data and an instruction to output a question whose answer exists in the model sentence data into a language model with the answer in the sentence. <How snippet display information is processed> (Appendix 34) a search query receiving step of receiving a search query; a question and answer data acquisition step of acquiring question and answer data related to the search query from a question and answer database; a search result acquisition step of acquiring search results from a search server using the search query; a transmitting step of transmitting the search results and the question and answer data; the question and answer database includes a plurality of question and answer data; An information processing method executed on a computer, wherein the question and answer data is data linking a question obtained by inputting model sentence data and an instruction to output a question whose answer exists in the model sentence data into a language model with the answer in the sentence. (Appendix 35) 35. The method of any of claims 19 to 34, wherein the language model is a large-scale language model. (Appendix 36) An information processing method, including the information processing method according to Supplementary Note 33 or 34 and the information processing method according to any one of Supplementary Notes 24 to 32. <Paywall screen display program> (Appendix 37) On the computer, A receiving process for receiving identification information of document data; A text data acquisition process for acquiring text data associated with the identification information of the text data from a text database; a question and answer data acquisition process for acquiring question and answer data linked to the identification information of the text data from a question and answer database; a transmission process for transmitting the text data and the question and answer data; the question and answer database includes a plurality of question and answer data; The question and answer data is data that links a question obtained by inputting model sentence data and an instruction to output a question whose answer exists in the model sentence data into a language model with the answer in the sentence. (Appendix 38) The transmission process transmits a part of the text data and purchase necessity data of the text data, a purchase instruction determination process for determining whether or not a purchase instruction is included in the text data; The program described in Appendix 37, further comprising a text data transmission process for transmitting the remainder or all of the text data when it is determined that the purchase instruction has been received. (Appendix 39) 39. The information processing device according to claim 37 or 38, wherein the sending process includes display instruction data for instructing display of a part of the text data and the question and answer data. <Natural language query program> (Appendix 40) On the computer, A search query receiving process for receiving a search query; a question and answer data acquisition process for acquiring question and answer data related to the search query from a question and answer database; a transmission process for transmitting the question and answer data; the question and answer database includes a plurality of question and answer data; The question and answer data is data that links a question obtained by inputting model sentence data and an instruction to output a question whose answer exists in the model sentence data into a language model with the answer in the sentence. (Appendix 41) The question and answer data is linked to an index, The program according to claim 40, wherein the question and answer data acquisition process performs an index search using the search query and an index linked to the question and answer data, and acquires question and answer data related to the search query from the question and answer database. <Q&Aデータベースのプログラム> (Appendix 42) On the computer, A text data acquisition process for acquiring text; an input process for inputting the text data and an instruction to output a question whose answer is present in the text data into a language model; a program for causing the program to execute a storage process for storing a pair of the question generated from the language model and the answer in a question and answer database as question and answer data. (Appendix 43) A type determination process is included for determining the type of the text data. The program described in Appendix 42, wherein the input processing inputs the text data and an instruction to output a question whose answer is present in the text data, which is set according to the type of the text data, into the language model. (Appendix 44) a number setting process for setting the number of questions to be output by the language model based on the number of characters of the text data; the input process inputs the text data, an instruction to output a set of questions and answers whose answers are present in the text data, and the number of questions to be output to a language model; The method further includes a question determination process for determining whether the question generated from the language model satisfies a pre-specified evaluation criterion, 44. The program according to claim 42 or 43, wherein the storing process stores a pair of a question and an answer determined to satisfy the evaluation criterion in the question and answer database as the question and answer data. (Appendix 45) 45. The program according to any one of appendices 42 to 44, wherein the input processing inputs the text data, an instruction to output a question whose answer is present in the text data, and positive examples and / or negative examples of the question into the language model. (Appendix 46) 46. ​​The program according to any one of appendices 42 to 45, further comprising an index assignment process for assigning an index to the question and answer pair. (Appendix 47) The input process inputs the text data and an instruction to output a set of questions and answers in which answers exist in the text data to a language model, The storage process stores, as the question-and-answer data, the set of questions and answers generated from the language model in the question-and-answer database, which is a program according to any one of Appendices 42 to 46. (Appendix 48) The extraction process includes extracting, from the text data, an answer to a question generated from the language model. The storage process associates the question generated from the language model with the extracted answer, and stores them in the question-and-answer database as the question-and-answer data, which is a program according to any one of Appendices 42 to 47. (Appendix 49) A program including a program according to any one of Appendices 37 to 39 and a program according to any one of Appendices 42 to 48. (Appendix 50) A program including a program according to Appendix 40 or 41 and a program according to any one of Appendices 42 to 48. <Q&A display program> (Appendix 51) Cause a computer to Execute a reception process for receiving identification information of text data, A text data acquisition process for acquiring text data associated with the identification information of the text data from a text database, A question-and-answer data acquisition process for acquiring question-and-answer data related to the text data from a question-and-answer database, And execute a transmission process for transmitting the text data and the question-and-answer data, The question-and-answer database includes a plurality of question-and-answer data, The question-and-answer data is data in which a question obtained by inputting the model text data and an instruction to output a question in which an answer exists in the model text data to a language model is associated with the answer in the text, which is a program. <Snippet display program> (Appendix 52) On the computer, A search query receiving process for receiving a search query; a question and answer data acquisition process for acquiring question and answer data related to the search query from a question and answer database; a search result acquisition process for acquiring search results from a search server using the search query; a transmission process for transmitting the search results and the question and answer data; the question and answer database includes a plurality of question and answer data; The question and answer data is data that links a question obtained by inputting model sentence data and an instruction to output a question whose answer exists in the model sentence data into a language model with the answer in the sentence. (Appendix 53) 53. The program of any of claims 37 to 52, wherein the language model is a large-scale language model. (Appendix 54) A program comprising the program of appendix 51 or 52 and the program of any of appendices 42 to 48. [Industrial Applicability]

[0101] As described above, according to the present disclosure, it is possible to provide an information processing device capable of generating a database for storing questions whose answers exist in text data. Therefore, the present disclosure is extremely useful in, for example, online sales promotion fields and online search fields. [Explanation of symbols]

[0102] 1, 1A, 1B, 1C, 6, 7, 8, 9 Information processing device 11 Acquisition Department 12 Input section 13 Storage area 14 Extraction part 15 Type discrimination section 16 Number setting section 17 Question Judgment Department 2. Language Model 3 Q&A DB 3A Text DB 4. Telecommunications Network 5. User terminal 61, 71 Receiver 62, 72 Text data acquisition section 63, 73, 82, 92 Question and Answer Data Acquisition Section 64, 74, 83, 94 Transmitter 65 Purchase Instruction Judgment Department 66 Text data transmission unit 81, 91 Query receiving section 93 Search Result Acquisition Unit 100, 200, 300, 400, 600, 700, 800, 900 Information Processing Systems

Claims

1. A receiving unit for receiving identification information of document data; a text data acquisition unit that acquires text data associated with the identification information of the text data from a text database; a question and answer data acquisition unit that acquires question and answer data linked to identification information of the text data from a question and answer database; a transmission unit for transmitting the text data and the question and answer data, the question and answer database includes a plurality of question and answer data; The question and answer data is data linking a question obtained by inputting model sentence data and an instruction to output a question whose answer exists in the model sentence data into a language model with the answer in the sentence, in an information processing device.

2. The transmission unit transmits a part of the text data and purchase necessity data of the text data, a purchase instruction determination unit that determines whether or not a purchase instruction is provided for the text data; a text data transmission unit that transmits the remainder or all of the text data when it is determined that the purchase instruction has been received. The information processing device according to claim 1 .

3. The information processing apparatus according to claim 1 , wherein the transmission unit transmits a part of the text data and display instruction data for instructing a display of the question and answer data.

4. a search query receiving unit for receiving a search query; a question and answer data acquisition unit that acquires question and answer data related to the search query from a question and answer database; a transmission unit for transmitting the question and answer data; the question and answer database includes a plurality of question and answer data; the question and answer data is data in which a question obtained by inputting model sentence data and an instruction to output a question whose answer is present in the model sentence data into a language model is associated with the answer in the sentence, An information processing device, wherein the text data is a newspaper article, a magazine article, a book, an official document, a web medium, a report, a financial statement summary, a securities report, or a press release.

5. The question and answer data is linked to an index, The information processing device according to claim 4 , wherein the question and answer data acquiring unit performs an index search using the search query and an index linked to the question and answer data, and acquires the question and answer data related to the search query from the question and answer database.

6. A text data acquisition unit that acquires text data; an input unit that inputs the text data and an instruction to output a question whose answer is present in the text data into a language model; a storage unit that stores a set of the question generated from the language model and the answer in a question and answer database as question and answer data; a type determination unit that determines a type of the text data; The input unit inputs, to the language model, the text data and an instruction to output a question whose answer is present in the text data, the instruction being set according to a type of the text data.

7. A text data acquisition unit that acquires text data; an input unit that inputs the text data and an instruction to output a question whose answer is present in the text data into a language model; a storage unit that stores a set of the question generated from the language model and the answer in a question and answer database as question and answer data; a number setting unit that sets the number of questions to be output by the language model based on the number of characters of the text data; the input unit inputs the text data, an instruction to output a set of questions and answers whose answers are present in the text data, and the number of questions to be output to a language model; The system further includes a question determination unit that determines whether a question generated from the language model satisfies a pre-specified evaluation criterion, The storage unit stores the question and answer pair determined to satisfy the evaluation criterion in the question and answer database as the question and answer data.

8. The information processing device according to claim 6 , wherein the input unit inputs the text data, an instruction to output a question whose answer is present in the text data, and a positive example and / or a negative example of the question to the language model.

9. The information processing apparatus according to claim 6 , further comprising an index assigning unit that assigns an index to the pair of question and answer.

10. the input unit inputs the text data and an instruction to output a question and answer pair, the question and answer pair having an answer present in the text data, to a language model; The information processing apparatus according to claim 6 , wherein the storage unit stores a pair of a question and an answer generated from the language model in the question and answer database as the question and answer data.

11. an extraction unit that extracts answers to questions generated from the language model from the text data; The information processing apparatus according to claim 6 , wherein the storage unit associates the question generated from the language model with the extracted answer, and stores the associated question and answer data in the question and answer database.

12. An information processing device according to claim 1 or 2; A text data acquisition unit that acquires text data; an input unit that inputs the text data and an instruction to output a question whose answer is present in the text data into a language model; an information processing system including an information processing device, the information processing device including a storage unit that stores pairs of the questions generated from the language model and the answers in a question and answer database as question and answer data;

13. An information processing device according to claim 4 or 5; A text data acquisition unit that acquires text data; an input unit that inputs the text data and an instruction to output a question whose answer is present in the text data into a language model; an information processing system including an information processing device, the information processing device including a storage unit that stores pairs of the questions generated from the language model and the answers in a question and answer database as question and answer data;

14. A receiving unit for receiving identification information of document data; a text data acquisition unit that acquires text data associated with the identification information of the text data from a text database; a question and answer data acquisition unit that acquires question and answer data related to the text data from a question and answer database; a transmission unit for transmitting the text data and the question and answer data, the question and answer database includes a plurality of question and answer data; The question and answer data is data linking a question obtained by inputting model sentence data and an instruction to output a question whose answer exists in the model sentence data into a language model with the answer in the sentence, in an information processing device.

15. a search query receiving unit for receiving a search query; a question and answer data acquisition unit that acquires question and answer data related to the search query from a question and answer database; a search result acquisition unit that acquires search results from a search server using the search query; a transmission unit for transmitting the search results and the question and answer data; the question and answer database includes a plurality of question and answer data; The question and answer data is data linking a question obtained by inputting model sentence data and an instruction to output a question whose answer exists in the model sentence data into a language model with the answer in the sentence, in an information processing device.

16. The information processing device of claim 1 , wherein the language model is a large-scale language model.

17. a receiving step of receiving identification information of document data; a text data acquisition step of acquiring text data linked to the identification information of the text data from a text database; a question and answer data acquisition step of acquiring question and answer data linked to the identification information of the text data from a question and answer database; a transmitting step of transmitting the text data and the question and answer data; the question and answer database includes a plurality of question and answer data; An information processing method executed on a computer, wherein the question and answer data is data linking a question obtained by inputting model sentence data and an instruction to output a question whose answer exists in the model sentence data into a language model with the answer in the sentence.

18. a search query receiving step of receiving a search query; a question and answer data acquisition step of acquiring question and answer data related to the search query from a question and answer database; a transmitting step of transmitting the question and answer data; the question and answer database includes a plurality of question and answer data; the question and answer data is data in which a question obtained by inputting model sentence data and an instruction to output a question whose answer is present in the model sentence data into a language model is associated with the answer in the sentence, An information processing method executed on a computer, wherein the text data is a newspaper article, a magazine article, a book, an official document, a web medium, a report, a financial statement summary, a securities report, or a press release.

19. A text data acquisition step of acquiring text data; an input step of inputting the text data and an instruction to output a question whose answer is present in the text data into a language model; a storing step of storing a set of the question generated from the language model and the answer in a question and answer database as question and answer data; a type determination step of determining a type of the text data, An information processing method executed on a computer, in which the input process inputs the text data and an instruction to output a question whose answer exists in the text data, which is set according to the type of the text data, into the language model.

20. A text data acquisition step of acquiring text data; an input step of inputting text data and an instruction to output a question whose answer is present in the text data into a language model; a storing step of storing a set of the question generated from the language model and the answer in a question and answer database as question and answer data; A number setting method for setting the number of questions to be output by the language model based on the number of characters in the text data. and In the input step, the text data, an instruction to output a set of questions and answers whose answers are present in the text data, and the number of questions to be output are input to a language model; Furthermore, the method determines whether the question generated from the language model satisfies a pre-specified evaluation criterion. A question determination step is included, A computer-executed information processing method, wherein in the storing step, pairs of questions and answers determined to satisfy the evaluation criterion are stored in the question and answer database as the question and answer data.

21. a receiving step of receiving identification information of document data; a text data acquisition step of acquiring text data linked to the identification information of the text data from a text database; a question and answer data acquisition step of acquiring question and answer data related to the text data from a question and answer database; a transmitting step of transmitting the text data and the question and answer data; the question and answer database includes a plurality of question and answer data; An information processing method executed on a computer, wherein the question and answer data is data linking a question obtained by inputting model sentence data and an instruction to output a question whose answer exists in the model sentence data into a language model with the answer in the sentence.

22. a search query receiving step of receiving a search query; a question and answer data acquisition step of acquiring question and answer data related to the search query from a question and answer database; a search result acquisition step of acquiring search results from a search server using the search query; a transmitting step of transmitting the search results and the question and answer data; the question and answer database includes a plurality of question and answer data; An information processing method executed on a computer, wherein the question and answer data is data linking a question obtained by inputting model sentence data and an instruction to output a question whose answer exists in the model sentence data into a language model with the answer in the sentence.

23. On the computer, A receiving process for receiving identification information of document data; A text data acquisition process for acquiring text data associated with the identification information of the text data from a text database; a question and answer data acquisition process for acquiring question and answer data linked to the identification information of the text data from a question and answer database; a transmission process for transmitting the text data and the question and answer data; the question and answer database includes a plurality of question and answer data; The question and answer data is data that links a question obtained by inputting model sentence data and an instruction to output a question whose answer exists in the model sentence data into a language model with the answer in the sentence.

24. On the computer, A search query receiving process for receiving a search query; a question and answer data acquisition process for acquiring question and answer data related to the search query from a question and answer database; a transmission process for transmitting the question and answer data; the question and answer database includes a plurality of question and answer data; the question and answer data is data in which a question obtained by inputting model sentence data and an instruction to output a question whose answer is present in the model sentence data into a language model is associated with the answer in the sentence, The program, wherein the text data is a newspaper article, a magazine article, a book, an official document, a web medium, a report, a financial statement summary, a securities report, or a press release.

25. On the computer, A text data acquisition process for acquiring text data; an input process for inputting the text data and an instruction to output a question whose answer is present in the text data into a language model; a storage process of storing a set of the question generated from the language model and the answer in a question and answer database as question and answer data; and executing a type determination process for determining the type of the document data. The input process inputs, to the language model, the text data and an instruction to output a question whose answer is present in the text data, the instruction being set according to the type of the text data.

26. A computer comprising: A text data acquisition process for acquiring text data; an input process for inputting text data and an instruction to output a question whose answer is present in the text data into a language model; a storage process of storing a set of the question generated from the language model and the answer in a question and answer database as question and answer data; a number setting process for setting the number of questions to be output by the language model based on the number of characters of the text data; In the input process, the text data, an instruction to output a set of questions and answers whose answers are present in the text data, and the number of questions to be output are input to a language model; Furthermore, the method determines whether the question generated from the language model satisfies a pre-specified evaluation criterion. Execute a question determination process, The storage process stores the question and answer pairs determined to satisfy the evaluation criterion in the question and answer database as the question and answer data.

27. On the computer, A receiving process for receiving identification information of document data; A text data acquisition process for acquiring text data associated with the identification information of the text data from a text database; a question and answer data acquisition process for acquiring question and answer data related to the text data from a question and answer database; a transmission process for transmitting the text data and the question and answer data; the question and answer database includes a plurality of question and answer data; The question and answer data is data that links a question obtained by inputting model sentence data and an instruction to output a question whose answer exists in the model sentence data into a language model with the answer in the sentence.

28. On the computer, A search query receiving process for receiving a search query; a question and answer data acquisition process for acquiring question and answer data related to the search query from a question and answer database; a search result acquisition process for acquiring search results from a search server using the search query; a transmission process for transmitting the search results and the question and answer data; the question and answer database includes a plurality of question and answer data; The question and answer data is data that links a question obtained by inputting model sentence data and an instruction to output a question whose answer exists in the model sentence data into a language model with the answer in the sentence.

Images