Information processing system, information processing method, and program

Abstract

To effectively generate an answer corresponding to input data.SOLUTION: An information processing system includes an accumulated data storage unit configured to store accumulated data, a provider storage unit configured to store information for specifying a provider of the accumulated data, a search unit configured to search the accumulated data related to input data, a generation unit configured to generate an answer by giving a prompt including the searched accumulated data and the input data to a large-scale language model, and a reward giving unit configured to specify the provider corresponding to the searched accumulated data from the provider storage unit and give a reward to the specified provider.SELECTED DRAWING: Figure 1

Description

[Technical Field]

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

[0002] Technologies that allow machines to automatically respond to questions and requests from users have been developed, and are now being provided as automatic response services such as bots (see Patent Document 1). [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Laid-Open No. 2009-3533 Summary of the Invention [Problem to be solved by the invention]

[0004] Retrieval-Augmented Generation (RAG), which combines text generation processing using large-scale language models (LLMs) with external information retrieval to generate answers, is beginning to be used, but it does not provide compensation to providers of the external information used by RAG.

[0005] The present invention has been made in view of the above background, and has an object to provide a technique that can effectively generate an answer according to input data. [Means for solving the problem]

[0006] The main invention of the present invention for solving the above problem is an information processing system comprising: an accumulated data storage unit that stores accumulated data; a provider storage unit that stores information identifying a provider of the accumulated data; a search unit that searches the accumulated data related to input data; a generation unit that provides a prompt including the searched accumulated data and the input data to a large-scale language model to generate an answer; and a reward granting unit that identifies the provider corresponding to the searched accumulated data from the provider storage unit and grants a reward to the identified provider.

[0007] Other problems and solutions disclosed in this application will be made clear in the section on preferred embodiments of the invention and the drawings. [Effects of the Invention]

[0008] According to the present invention, it is possible to effectively generate a response according to input data. [Brief explanation of the drawings]

[0009] [Figure 1] FIG. 1 is a diagram illustrating an example of the overall configuration of an information processing system. [Figure 2] FIG. 2 illustrates an example of a hardware configuration of a management server 2. [Figure 3] FIG. 2 illustrates an example of the software configuration of a management server 2. [Figure 4] FIG. 10 is a diagram illustrating the operation of the management server 2. DETAILED DESCRIPTION OF THE INVENTION

[0010] <System Overview> An information processing system according to one embodiment of the present invention will be described below. The information processing system of this embodiment aims to provide information through RAG, and to motivate providers of external information to provide their accumulated data by providing rewards in accordance with citations in the RAG, thereby ensuring that answers are provided effectively through the RAG.

[0011] 1 is a diagram showing an example of the overall configuration of an information processing system. The information processing system of this embodiment is configured to include a management server 2. The management server 2 is communicably connected to a user terminal 1 via a communication network. The communication network is, for example, the Internet, and is constructed using a public telephone line network, a mobile phone line network, a wireless communication path, Ethernet (registered trademark), etc.

[0012] The user terminal 1 is a computer operated by a user, and may be, for example, a smartphone, a tablet computer, or a personal computer.

[0013] The management server 2 is a computer that provides information via the RAG. The management server 2 may be a general-purpose computer such as a workstation or personal computer, or may be logically realized by cloud computing.

[0014] <Administration Server> FIG. 2 is a diagram illustrating an example of the hardware configuration of the management server 2. Note that the illustrated configuration is an example, and other configurations may also be used. The management server 2 includes a CPU 201, a memory 202, a storage device 203, a communication interface 204, an input device 205, and an output device 206. The storage device 203 stores various data and programs, and is, for example, a hard disk drive, a solid state drive, or a flash memory. The communication interface 204 is an interface for connecting to a communication network, and is, for example, an adapter for connecting to Ethernet (registered trademark), a modem for connecting to a public telephone network, a wireless communication device for wireless communication, or a USB (Universal Serial Bus) connector or an RS232C connector for serial communication. The input device 205 is used to input data, and is, for example, a keyboard, a mouse, a touch panel, a button, a microphone, or the like. The output device 206 is used to output data, and is, for example, a display, a printer, a speaker, or the like. Each functional unit of the management server 2 described below is realized by the CPU 201 reading a program stored in the storage device 203 into the memory 202 and executing it, and each storage unit of the management server 2 is realized as part of the storage area provided by the memory 202 and the storage device 203.

[0015] 3 is a diagram illustrating an example of the software configuration of the management server 2. The management server 2 includes an accumulated data storage unit 231, a provider storage unit 232, a search unit 211, a generation unit 212, a reward granting unit 213, and an accumulated data registration unit 214.

[0016] <Storage section>

[0017] The accumulated data storage unit 231 stores accumulated data. The accumulated data includes, for example, text data, image data, and audio data transmitted from the user terminal 1. Specific examples of text data include questions entered by a user, sentences extracted from web pages, and portions of books and papers. Specific examples of image data include photos taken by a user, images extracted from web pages, and diagrams included in books and papers. Specific examples of audio data include audio recorded by a user, audio extracted from videos, and podcasts. The accumulated data may also be structured data, such as tabular data, XML data, and JSON data. In this embodiment, the accumulated data storage unit 231 is assumed to be a vector store that stores vector information in which all or part of the accumulated data has been embedded. The accumulated data storage unit 231 can store information identifying accumulated data (such as text data) and the vector information in which all or part of the accumulated data has been embedded, in association with each other.

[0018] The provider storage unit 232 stores information that identifies the provider of the accumulated data. The information that identifies the provider is, for example, the user ID of the user terminal 1 or identification information of the user terminal 1. In this embodiment, the information that identifies the provider is the user ID. The user ID is, for example, a character string used when the user of the user terminal 1 logs in to the system. Alternatively, the user ID may be the email address of the user of the user terminal 1. The information that identifies the provider may also be identification information of the user terminal 1, such as the MAC address, IMEI (International Mobile Equipment Identifier), or IP address of the user terminal 1. The provider storage unit 232 stores information that identifies the accumulated data using the user ID as a key. The information that identifies the accumulated data is, for example, a file path, a database key, a URL, or the like.

[0019] <Functional section> The search unit 211 searches for stored data related to input data. The input data is, for example, a question sent from the user terminal 1. The search unit 211 can search the stored data storage unit 231 for stored data similar to the input data. Specifically, the search unit 211 calculates the similarity between the input data and each stored data, and outputs stored data whose similarity is equal to or greater than a threshold as a search result. The similarity is calculated using, for example, an edit distance, cosine similarity, or Jaccard similarity between the input data and the stored data. The edit distance is an index representing the number of edit operations required to convert one character string into another character string. The cosine similarity is an index representing the cosine of the angle between two vectors. The Jaccard similarity is an index representing the value obtained by dividing the number of elements in the intersection of two sets by the number of elements in the union. The search unit 211 may also calculate the similarity between the input data and each stored data using a machine learning model. The machine learning model is, for example, a neural network model using deep learning, and can learn a method for calculating similarity from a large amount of pairs of input data and accumulated data. In this embodiment, it is assumed that the search unit 211 creates vector information by embedding all or part of the input data, and searches for accumulated data similar to the input data according to the distance between the created vector information and the vector information stored in the accumulated data storage unit 231.

[0020] The generation unit 212 generates an answer by providing a prompt including the stored data and input data searched by the search unit 211 to the large-scale language model. Specifically, the generation unit 212 generates a prompt by separating and concatenating the stored data and input data searched by the search unit 211 with a predetermined symbol, and inputs the generated prompt into the large-scale language model to generate an answer to the input data. The generation unit 212 may also generate an answer for each of the stored data searched by the search unit 211 and integrate the obtained answers to generate a final answer. The integration of answers is achieved, for example, by extracting the most frequently occurring sentence or phrase from the multiple answers or by summarizing the multiple answers. The large-scale language model used may be, for example, a general-purpose language model such as GPT-3, PaLM, OPT, or Chinese-LLaMa, or may be a language model specialized for question answering. In this embodiment, the large-scale language model is assumed to generate text, but it may also be a multimodal model capable of generating output other than text. For example, the large-scale language model may be a diffusion model, and may generate images, sounds, etc. according to specifications in the input data, the format of the stored data, etc. The generation unit 212 can transmit to the user terminal 1 the answer generated by the large-scale language model as is, or create a message based on the generated answer. The generation unit 212 can provide (transmit to the user terminal 1) information (e.g., title, URL, etc.) identifying the stored data used to generate the answer. The stored data used to generate the answer may be stored data included in a prompt. Furthermore, the generation unit 212 may include an instruction indicating the stored data used to generate the answer in the prompt, causing the large-scale language model to respond with the stored data used to generate the answer. The generation unit 212 may read out the provider of the stored data used to generate the answer from the provider storage unit 232, and provide (transmit to the user terminal 1) information indicating the read out provider (e.g., username, nickname, etc.).

[0021] The reward granting unit 213 identifies providers corresponding to the stored data searched by the search unit 211 from the provider storage unit 231 and grants rewards to all or some of the identified providers. For example, the reward granting unit 213 acquires the ID of the stored data searched by the search unit 211 and identifies the ID of the user who provided the stored data by referring to the provider storage unit 231. The reward granting unit 213 grants a reward to a reward account linked to the identified user ID. The reward may be, for example, points circulating in the market, virtual currency, or a token using blockchain technology. The reward may also be a coupon or the like. For example, the reward may be digital content or a physical lottery ticket, and the number of lottery attempts or the probability of winning the lottery may be the amount of the reward. The reward granting unit 213 may select providers to whom a reward will be given by lottery or randomly from among the providers of the searched stored data, and grant a reward only to the selected provider. The reward granting unit 213 may grant a reward only to the provider of the stored data used to generate the answer.

[0022] The reward granting unit 213 can determine the size of the reward depending on the number of times the stored data was used to generate an answer. The reward granting unit 213 can determine the size of the reward depending on the similarity of the stored data to the input data. The reward granting unit 213 can determine the size of the reward depending on the similarity between the answer and the stored data included in the prompt.

[0023] The accumulated data registration unit 214 registers accumulated data received from the provider's user terminal 1 in the accumulated data storage unit 231. The accumulated data registration unit 214 registers accumulated data in the accumulated data storage unit 231, and can also register information identifying the accumulated data and information identifying the provider in the provider storage unit 232 in association with each other. The accumulated data registration unit 214 can register all of the accumulated data in the accumulated data storage unit 231 as vector information by embedding processing. Here, embedding processing is a type of preprocessing in natural language processing, and is a process of converting words and sentences into fixed-length vectors. In addition, in order to efficiently process large amounts of accumulated data and speed up searches, the accumulated data registration unit 214 can divide the accumulated data into several blocks (chunk data), perform embedding processing for each of the divided chunk data, and register vector information in the accumulated data storage unit 231.

[0024] <Operation> FIG. 4 is a diagram illustrating the operation of the management server 2.

[0025] The management server 2 receives input data from the user terminal 1 (S301), searches for stored data similar to the received input data (S302), provides a prompt including the input data and stored data to a large-scale language model to generate an answer (S303), and sends the generated answer to the user terminal 1 (S304).

[0026] The management server 2 identifies a provider corresponding to stored data similar to the searched input data (S305), and provides a reward to the identified provider (S306).

[0027] As described above, according to the information processing system of this embodiment, a reward can be given to the provider of the stored data used by the RAG.

[0028] Although the present embodiment has been described above, the above embodiment is intended to facilitate understanding of the present invention and is not intended to limit the present invention. The present invention may be modified or improved without departing from the spirit thereof, and equivalents thereof are also included in the present invention.

[0029] For example, the processing by each of the functional units of the management server 2 described above may be performed by any of the functional units. Also, a different functional unit that performs part of the processing by each of the functional units described above may be added. Also, the functional units of the management server 2 may be distributed across multiple computers.

[0030] Furthermore, the information stored in each storage unit of the management server may be stored in any of the storage units. That is, the information stored in the above-mentioned multiple storage units may be stored in one storage unit, or part of the information stored in one of the above-mentioned storage units may be stored in another storage unit.

[0031] <Variation 1> In the above embodiment, the reward granting unit 213 grants a reward based on the number of times the stored data was used to generate an answer or the degree of similarity, but this is not limited to this. For example, the reward granting unit 213 may grant a reward based on the size, format, creation date, etc. of the stored data. Specifically, the reward granting unit 213 may grant a reward for stored data of a predetermined size or larger, or for stored data in a predetermined format (e.g., structured data, image data, etc.). Furthermore, the reward granting unit 213 may grant a reward for stored data created within a predetermined period (e.g., the past week, etc.). This can encourage the provision of large data, data in a format that is easy to process, and the latest data.

[0032] The reward granting unit 213 may also grant a reward according to the quality of the stored data. The quality may be evaluated manually or automatically by AI, for example. In the case of manual evaluation, the reward granting unit 213 may have an operator evaluate the quality of the stored data based on pre-prepared quality evaluation criteria. The quality evaluation criteria may be, for example, a checklist including items such as the accuracy of information contained in the stored data, the comprehensiveness of the stored data, and the usefulness of the stored data. Alternatively, a subjective evaluation by a third party may be performed without setting quality evaluation criteria. For example, when generating an answer using the stored data, feedback from the user on the stored data (e.g., an evaluation value from 0 to 5) may be accepted, and the stored data may be scored (points may be deducted, in which case a negative number may be included in the evaluation value, e.g., an evaluation value from -5 to +5). For example, the average value of the scored stored data may be used as the quality. The reward granting unit 213 may determine the amount of reward based on the results of the quality evaluation. In the case of automatic evaluation using AI, the reward granting unit 213 automatically evaluates the quality of the stored data based on quality evaluation criteria that have been trained in a machine learning model in advance. The machine learning model can predict the quality from the features of the stored data using a large amount of stored data that has been manually quality-evaluated as training data. The reward granting unit 213 can then determine the amount of reward based on the results of the quality evaluation by the machine learning model.

[0033] Furthermore, the reward granting unit 213 may grant a reward according to the degree of contribution of the provider of stored data. The degree of contribution of the provider may be calculated based on, for example, the cumulative number of pieces of stored data provided by the provider, the cumulative size of the stored data provided by the provider, the average quality of the stored data provided by the provider, etc. Then, the reward granting unit 213 can grant a reward in an amount proportional to the degree of contribution of the provider. This makes it possible to give preferential treatment to users who continue to provide high-quality stored data over the long term.

[0034] <Variation 2> The reward granting unit 213 may also determine the amount of reward by combining multiple perspectives. For example, the reward granting unit 213 may determine the amount of reward based on the number of times the accumulated data was used to generate an answer, the size of the accumulated data, and the result of the quality evaluation of the accumulated data. Specifically, the reward granting unit 213 may calculate the amount of reward according to the following formula: R=β1·N+β2·S+α R=β1·N+β3·Q+α R=β1·N+β2·S+β3·Q+α

[0035] Here, R is the amount of reward, α is a constant parameter, β1 and β2 are coefficients, N is the number of stored data used to generate an answer, S is the size of the stored data, and Q is the result of the quality evaluation of the stored data. Note that the above formula is an example, and the amount of reward may be calculated using other functions.

[0036] Furthermore, the reward granting unit 213 may calculate the amount of reward using different functions depending on the format of the stored data. For example, the reward granting unit 213 may calculate the amount of reward using different formulas for text data and image data, thereby changing the type of reward.

[0037] Furthermore, the reward granting unit 213 may change the amount of reward depending on the creation date and time of the accumulated data. For example, the reward granting unit 213 may calculate the amount of reward according to the following formula. R=β1·N-β4·(T-T0)+α R=β1·N+β2·S-β4·(T-T0)+α R=β1·N+β3·Q-β4·(T-T0)+α R=β1·N+β2·S+β3·Q-β4·(T-T0)+α

[0038] Here, R is the amount of reward, α is a constant parameter, N is the number of times stored data was used to generate an answer, S is the size of the stored data, Q is the result of the quality evaluation of the stored data, T is the current date and time, and T0 is the date and time the stored data was created. Note that the above formula is just an example, and other functions can be used to calculate the amount of reward. This gradually reduces the reward for older data, allowing the most recent data to be collected preferentially.

[0039] <Variation 3> Furthermore, in the above embodiment, the reward granting unit 213 grants virtual currency, coupons, lottery tickets, and the like to the provider, but the present invention is not limited to this. For example, the reward granting unit 213 may grant a badge or status indicating the provider's recognition or trustworthiness. Specifically, the reward granting unit 213 may grant a rank such as "silver," "gold," or "platinum" according to the provider's level of contribution. The rank is determined based on, for example, the cumulative number of accumulated data provided by the provider or the average quality of the accumulated data provided by the provider. Then, the reward granting unit 213 displays a badge corresponding to the provider's rank on the provider's profile page or on the accumulated data provided by the provider. This can increase the provider's reputation and motivate the provider.

[0040] Furthermore, the reward granting unit 213 may grant a special status to a provider according to the amount of accumulated reward. For example, the reward granting unit 213 may grant a title such as "premium provider" to a provider whose accumulated amount of reward exceeds a predetermined threshold. The reward granting unit 213 may then provide special services to premium providers. Examples of special services include a dedicated inquiry desk for premium providers and preferential display of accumulated data provided by premium providers. This can increase the satisfaction of providers and increase their motivation.

[0041] Furthermore, the reward granting unit 213 may create a ranking of providers to encourage competition among providers. For example, the reward granting unit 213 may create a ranking of providers based on the cumulative amount of rewards received by the providers or the degree of contribution of the providers. Then, the reward granting unit 213 may grant special rewards to top-ranked providers. Examples of special rewards include invitations to events that only top-ranked providers can participate in, and limited items that only top-ranked providers can use. This can encourage competition among providers, and is expected to result in the provision of higher quality accumulated data.

[0042] Furthermore, the reward granting unit 213 may visualize the activity history of the providers to increase their motivation. For example, the reward granting unit 213 may compile, for each provider, the number of accumulated data provided to date, the amount of reward earned to date, the number of accumulated data used to generate answers to date, and the like, and present the graphs to the providers. This allows the providers to realize the results of their activities, and is expected to encourage them to provide accumulated data more proactively.

[0043] <Variation 4> Furthermore, in the above embodiment, the accumulated data storage unit 231 stores the accumulated data received from the provider as is, but in order to protect the privacy of the provider, the accumulated data may be anonymized or encrypted before being stored. For example, the accumulated data storage unit 231 may store the accumulated data after deleting information that can identify the provider (such as name, address, and telephone number) from the accumulated data. Alternatively, the accumulated data storage unit 231 may encrypt the accumulated data before storing it, and store the decryption key in a highly secure location separate from the provider storage unit 232. This makes it possible to protect the privacy of the provider even if the accumulated data is leaked.

[0044] Furthermore, the accumulated data stored in the accumulated data storage unit 231 may be managed by the provider himself / herself. For example, the management server 2 may be provided with an interface that enables the provider to view, modify, delete, etc. the accumulated data. Specifically, the management server 2 provides a dedicated website that the provider can log in to, and after logging in, a list of the accumulated data provided by the provider is displayed on the screen. The provider can then select any accumulated data from the list and view, modify, or delete its contents. This will encourage providers to manage their own accumulated data more proactively, and it is expected that high-quality accumulated data will be provided.

[0045] Furthermore, the management server 2 may make the stored data provided by a provider public to other providers. For example, the management server 2 allows the provider to select whether to make his / her stored data public or private on a screen after logging in. Then, the management server 2 stores the stored data selected for public access in the stored data storage unit 231 in a state that can be viewed by other providers. This promotes the sharing of stored data between providers, and is expected to result in the provision of higher quality stored data.

[0046] Furthermore, the management server 2 may be equipped with an automatic classification function using AI to assist providers in managing their stored data. For example, the management server 2 automatically classifies the stored data provided by the provider using a machine learning model and presents the classification results to the provider. The machine learning model uses a large amount of stored data as training data and can predict the classification from the features of the stored data. This makes it easier for providers to organize their own stored data and manage it more efficiently.

[0047] <Variation 5> In the above embodiment, the search unit 211 performs a search based on the similarity between the input data and the stored data. However, this is not limiting. For example, the search unit 211 may perform a search by taking into account keywords included in the input data, a search intent estimated from the input data, the context of the input data, and the like. Specifically, the search unit 211 may divide the input data into keywords and search for stored data containing each keyword. The search unit 211 may also estimate a search intent from the input data and search for stored data that matches the search intent. The search intent may be, for example, "purchase a product," "make a restaurant reservation," or "how to solve a problem," and can be automatically estimated from the input data using a machine learning model. Furthermore, the search unit 211 may analyze the context surrounding the input data and search for stored data that matches the context. This allows for a search of more appropriate stored data.

[0048] Furthermore, the search unit 211 may create an index of the stored data to speed up searches. The index is, for example, a table that associates keywords included in the stored data with identifiers of the stored data that include those keywords. By using the index, the search unit 211 can quickly search for keywords included in the input data. Furthermore, the index may be created dynamically when the stored data is registered or updated, or may be created periodically by batch processing.

[0049] Furthermore, the search unit 211 may manage stored data in a distributed manner to speed up searches. For example, the search unit 211 may include multiple search nodes, each of which manages a portion of the stored data. The search unit 211 then distributes input data to multiple search nodes to perform searches. This allows for high-speed searches of large amounts of stored data. Search nodes may also be dynamically added or deleted to distribute the load.

[0050] Furthermore, the search unit 211 may use user feedback to improve search accuracy. For example, the search unit 211 may collect user satisfaction with search results and preferentially display search results with high satisfaction. Specifically, the search unit 211 can display a UI for inputting user satisfaction along with the search results and collect satisfaction from the user. Then, the search unit 211 can display search results with high satisfaction at the top of the list in subsequent searches. This makes it possible to provide search results that match the user's preferences.

[0051] Furthermore, the search unit 211 may analyze the user's search history to estimate the user's interests and recommend stored data that matches those interests. For example, the search unit 211 may analyze the user's search history to estimate the user's interests, such as keywords frequently searched by the user and trends in the stored data viewed by the user. The search unit 211 then actively recommends stored data that matches the estimated interests to the user. This makes it possible to provide information that matches the user's interests without the user having to search for it.

[0052] <Variation 6> Furthermore, the right to receive a reward for a provider can be made transferable. For example, when user A provides stored data, a user ID indicating user A and information identifying the stored data are associated and registered in the provider storage unit 232. Here, when the right to receive a reward is transferred from user A to user B (for example, by assignment), the user ID of user A corresponding to the information identifying the stored data can be updated to the user ID indicating user B.

[0053] Alternatively, a non-fungible token (NFT) representing the right to receive a reward may be issued. In this case, the management server 2 connects to a blockchain network and can mint an NFT linked to the accumulated data using a smart contract implemented on the blockchain network. Information identifying the accumulated data is linked to the minted NFT. Information identifying the accumulated data may be set in the metadata of the NFT. In this case, the management server 2 performs a process of paying a reward to the owner of the NFT linked to the accumulated data used to generate the answer. Alternatively, the NFT may be linked to information indicating the provider of the original accumulated data (e.g., the provider's ID or wallet), and a predetermined percentage of the reward may be paid to the original provider, with the remainder being paid to the owner of the NFT.

[0054] Furthermore, when an NFT is resold (secondary distribution), a portion of the transfer fee may be distributed to the provider of the original stored data. In this case, a transfer smart contract is deployed on a blockchain network, and when an NFT is transferred, the smart contract pays a portion of the transfer fee (for example, an arbitrary percentage such as 10%) to the provider of the stored data linked to the NFT to be transferred, and pays the remaining amount (after which system fees, gas fees, etc. can be deducted) to the transferor of the NFT. In this case, payment is assumed to be made in virtual currency.

[0055] <Disclosures> The present disclosure also includes the following configurations. [Item 1] an accumulated data storage unit that stores accumulated data; a provider storage unit that stores information identifying a provider of the stored data; a search unit that searches the stored data for input data; a reward granting unit that identifies the provider corresponding to the searched stored data from the provider storage unit and grants a reward to the identified provider; An information processing system comprising: [Item 2] 2. The information processing system according to claim 1, a generation unit that provides a prompt including all or a part of the searched stored data and the input data to a large-scale language model to generate an answer, and provides the generated answer; An information processing system characterized by: [Item 3] Item 2. The information processing system according to item 2, the reward granting unit determines the size of the reward according to the number of times the accumulated data has been used to generate the answer; An information processing system characterized by: [Item 4] Item 1, an information processing system according to item 1, the search unit searches for the stored data similar to the input data; the reward granting unit determines the magnitude of the reward according to the similarity of the stored data to the input data; An information processing system characterized by: [Item 5] Item 2. The information processing system according to item 2, the reward granting unit determines the size of the reward according to the degree of similarity between the answer and the stored data included in the prompt; An information processing system characterized by: [Item 6] storing the accumulated data in an accumulated data storage unit; storing information identifying a provider of the stored data in a provider storage unit; searching the stored data for input data; identifying the provider corresponding to the searched stored data from the provider storage unit and giving a reward to the identified provider; An information processing method characterized by being executed by a computer. [Item 7] storing the accumulated data in an accumulated data storage unit; storing information identifying a provider of the stored data in a provider storage unit; searching the stored data for input data; identifying the provider corresponding to the searched stored data from the provider storage unit and giving a reward to the identified provider; A program that causes a computer to execute the following. [Explanation of symbols]

[0056] 1. User terminal 2 Management Server

Claims

1. an accumulated data storage unit that stores accumulated data; a provider storage unit that stores information identifying a provider of the stored data; a search unit that searches the stored data for input data; a reward granting unit that identifies the provider corresponding to the searched stored data from the provider storage unit and grants a reward to all or part of the identified providers; An information processing system comprising:

2. 2. The information processing system according to claim 1, a generation unit that provides a prompt including all or a part of the searched stored data and the input data to a large-scale language model to generate an answer, and provides the generated answer; An information processing system characterized by:

3. 3. The information processing system according to claim 2, the reward granting unit determines the size of the reward according to the number of times the accumulated data has been used to generate the answer; An information processing system characterized by:

4. 2. The information processing system according to claim 1, the search unit searches for the stored data similar to the input data; the reward granting unit determines the magnitude of the reward according to the similarity of the stored data to the input data; An information processing system characterized by:

5. 3. The information processing system according to claim 2, the reward granting unit determines the size of the reward according to the degree of similarity between the answer and the stored data included in the prompt; An information processing system characterized by:

6. storing the accumulated data in an accumulated data storage unit; storing information identifying a provider of the stored data in a provider storage unit; searching the stored data for input data; Identifying the providers corresponding to the searched stored data from the provider storage unit, and providing rewards to all or some of the identified providers; An information processing method characterized by being executed by a computer.

7. storing the accumulated data in an accumulated data storage unit; storing information identifying a provider of the stored data in a provider storage unit; searching the stored data for input data; Identifying the providers corresponding to the searched stored data from the provider storage unit, and providing rewards to all or some of the identified providers; A program that causes a computer to execute the following.

Images