Information processing systems, information processing methods, and programs

Abstract

To enable the effective collection of templates. [Solution] An information processing system comprising: a template storage unit that stores prompt templates for a large-scale language model in association with template providers; a template selection unit that accepts the selection of one template from among the templates; a template provision unit that provides the selected template to the user; a prompt receiving unit that receives prompts created by the user based on the templates; an answer generation unit that provides the received prompts to the large-scale language model to generate answers; and a reward granting unit that identifies the provider of the template used to create the received prompt from the template storage unit and grants a reward to the identified provider.

Description

Technical Field

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

Background Art

[0002] Patent Document 1 discloses a technique for creating a structured document using a template.

Prior Art Document

Patent Document

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] It is necessary to widely collect templates for various uses.

[0005] The present invention has been made in view of such a background, and an object thereof is to provide a technique capable of effectively collecting templates.

Means for Solving the Problems

[0006] The main invention of the present invention for solving the above problems is an information processing system comprising: a template storage unit that stores prompt templates to be given to a large-scale language model in association with the provider of the templates; a template selection unit that accepts the selection of one template from among the templates; a template provision unit that provides the selected template to a user; a prompt receiving unit that receives a prompt created by the user based on the template; a response generation unit that provides the received prompt to the large-scale language model to generate a response; and a reward granting unit that identifies the provider of the template used to create the received prompt from the template storage unit and grants a reward to the identified provider.

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

[0008] According to the present invention, templates can be collected effectively. [Brief explanation of the drawing]

[0009] [Figure 1] This figure shows an example of the overall configuration of an information processing system. [Figure 2] This figure shows an example of the hardware configuration of management server 2. [Figure 3] This figure shows an example of the software configuration for management server 2. [Figure 4] This diagram illustrates the operation of management server 2. [Modes for carrying out the invention]

[0010] <System Overview> The following describes an information processing system according to one embodiment of the present invention. The information processing system of this embodiment attempts to generate answers using a large-scale language model (LLM). In the information processing system of this embodiment, prompt templates to be given to the LLM are prepared, and users can create prompts using these templates. Rewards are given to providers of templates that are used, thereby incentivizing the registration of templates.

[0011] Figure 1 shows an example of the overall configuration of an information processing system. The information processing system in this embodiment includes a management server 2. The management server 2 is connected to the user terminal 1 via a communication network. The communication network is, for example, the internet and is constructed using public telephone networks, mobile phone networks, wireless communication channels, Ethernet (registered trademark), etc.

[0012] User terminal 1 is an information processing device operated by a user. User terminal 1 can include various forms of devices, such as portable information processing devices and stationary information processing devices. These devices are equipped with a user interface and have the ability to communicate with management server 2 via a communication network. Examples of portable information processing devices include, but are not limited to, smartphones, tablet computers, laptop computers, and wearable devices. Examples of stationary information processing devices include, but are not limited to, desktop computers, smart TVs, and game consoles. New forms of information processing devices developed in the future can also function as user terminal 1 of the present invention. User terminal 1 is equipped with basic computing resources such as a processor, memory, storage, input / output interfaces, and communication interfaces. These resources enable user terminal 1 to process information in various formats such as text, images, and audio, enable interaction with the user, and realize communication with management server 2. The specific form and specifications of user terminal 1 may change in accordance with technological advancements and market needs, and the present invention can flexibly respond to these changes. The important thing is that user terminal 1 can receive user input and communicate with management server 2 to perform functions such as selecting templates and sending prompts.

[0013] The management server 2 may be a general-purpose computer such as a workstation or personal computer, or it may be logically implemented through cloud computing.

[0014] <Management Server> Figure 2 shows an example of the hardware configuration of the management server 2. Note that the illustrated configuration is just one example, and other configurations are also possible. The management server 2 includes a CPU 201, memory 202, storage device 203, communication interface 204, input device 205, and output device 206. The storage device 203 stores various data and programs, such as a hard disk drive, solid-state drive, or flash memory. The communication interface 204 is an interface for connecting to a communication network, such as an adapter for connecting to Ethernet®, a modem for connecting to a public telephone network, a wireless communication device for wireless communication, or a USB (Universal Serial Bus) connector or RS232C connector for serial communication. The input device 205 is for inputting data, such as a keyboard, mouse, touch panel, button, or microphone. The output device 206 is for outputting data, such as a display, printer, or speaker. Furthermore, each functional unit of the management server 2, as described later, is realized by the CPU 201 reading programs stored in the storage device 203 into memory 202 and executing them, and each storage unit of the management server 2 is realized as part of the storage area provided by memory 202 and storage device 203.

[0015] Figure 3 shows an example of the software configuration of the management server 2. The management server 2 includes a template storage unit 231, a template selection unit 211, a template provision unit 212, a prompt reception unit 213, an answer generation unit 214, a reward granting unit 215, a template reception unit 216, a template registration unit 217, and a template correction unit 218.

[0016] <Storage section> The template storage unit 231 stores templates for prompts given to the large language model. The template storage unit 231 stores templates in association with the providers of the templates. The template storage unit 231 can store templates, for example, in association with information identifying the provider (such as a user ID). The template storage unit 231 can store an explanation of the template in association with the template. When the registered template has been revised, the template storage unit 231 can store information identifying the provider who created the original template (original user ID) in association with the template.

[0017] <Functional unit> The template selection unit 211 accepts one selection from among the templates. The template selection unit 211 can create, for example, screen data for displaying a list of templates stored in the template storage unit 231, transmit it to the user terminal 1, display the list of templates on the user terminal 1, and allow the user to select a template from the list. The list of templates may be a list of explanations of the templates. The template selection unit 211 can accept, for example, an input of a keyword or key phrase from the user, search the template storage unit 231 for an explanation corresponding to the received keyword or key phrase, and display a list of templates corresponding to the searched explanation on the user terminal 1.

[0018] The template selection unit 211 can accept the selection of a template by any of the following methods, or a combination of these. Note that it is not limited to these methods.

[0019] (1) Selection by category: A method in which templates are classified into a plurality of categories, the user first selects a category, and then selects a template from within the selected category. (2) Keyword search: A method in which related templates are searched based on the keyword input by the user, and a selection is made from the search results. (3) Natural language requests: A method where the user inputs requests in natural language and the AI recommends the template most suitable for that request. (4) Recommendations based on the user's usage history: A method of analyzing the user's past template usage history and recommending appropriate templates. (5) Popularity ranking: A method of ranking templates based on the usage frequency and evaluations by other users, and preferentially displaying the top-ranked templates. (6) Tag-based selection: A method of narrowing down templates by assigning multiple tags to each template and allowing the user to select tags.

[0020] Also, the template selection unit 211 may recommend templates based on the user's profile. For example, the template selection unit 211 can consider information such as the user's past selection history, interests and concerns, occupation, etc., and perform template recommendations optimized for each user.

[0021] Also, the template selection unit 211 may recommend templates according to the context. For example, the template selection unit 211 can obtain context information such as the task the user is currently working on or the application used immediately before, and perform template recommendations according to the obtained context information.

[0022] The template providing unit 212 provides the selected template to the user. The template providing unit 212 can receive information identifying the template (e.g., template ID) from the user terminal 1, read out the template corresponding to the received template ID from the template storage unit 231, and transmit it to the user terminal 1.

[0023] The template providing unit 212 can also provide the content of the template step by step, for example. For example, in the case of a complex template, the template providing unit 212 can provide the content of the template step by step according to the user's understanding level and progress.

[0024] Furthermore, the template provider 212 may also display interactive guidance. For example, the template provider 212 can provide detailed explanations and usage examples for each part of the template using tooltips or pop-up windows.

[0025] Furthermore, the template provider 212 can provide customization options. For example, the template provider 212 can allow users to customize specific parts of the template.

[0026] The prompt receiving unit 213 receives prompts from the user. In this embodiment, it is assumed that the prompts received by the prompt receiving unit 213 are created based on templates provided by the template providing unit 212. For example, the template providing unit 212 sends screen data for creating an editable form screen to the user terminal 1, the user edits the template on the user terminal 1, and the user terminal 1 sends the edited result (prompt) to be received by the prompt receiving unit 213. The prompt receiving unit 213 can receive prompts along with information that identifies the template. In this case, information that identifies the template (e.g., a template ID) can be embedded in the form screen for template editing.

[0027] The prompt receiving unit 213 can receive prompts by any of the following methods, or a combination thereof. However, it is not limited to these methods.

[0028] (1) Text input: A method in which the user directly enters prompts into a text form. (2) Voice input: A method in which the user enters prompts by voice and converts them into text using speech recognition technology. (3) Template editing: A method of creating prompts by editing or replacing parts of the provided template. (4) Combining multiple templates: A method of creating a prompt by selecting multiple templates and combining them. (5) Step-by-step question answering: A method in which the system asks questions in a step-by-step manner and automatically generates prompts based on the user's answers. (6) Input of images and shapes: A method for generating prompts that include images and shapes entered by the user, along with their visual information.

[0029] Furthermore, the prompt receiving unit 213 may perform prompt validation. For example, the prompt receiving unit 213 can perform grammar checks and filter for prohibited words when a user enters a prompt.

[0030] Furthermore, the prompt receiving unit 213 may also manage the prompt history. In this case, it is possible to manage the history of prompts created by the user in the past and enable their reuse or editing.

[0031] Furthermore, the prompt receiving unit 213 may be configured to provide a collaborative editing function. For example, it may be possible to allow multiple users to edit a single prompt simultaneously.

[0032] The response generation unit 214 provides the received prompt to the large-scale language model to generate a response. The large-scale language model may be provided by the management server 2, or prompts may be provided to the large-scale language model by calling an API provided by an external server that provides the large-scale language model.

[0033] The response generation unit 214 can be configured to support multiple large-scale language models and can select the optimal model according to the content of the prompt and the characteristics of the required response.

[0034] Furthermore, the response generation unit 214 can generate responses in stages when the response is long or complex, and accept user approval or modification at each stage.

[0035] Furthermore, the response generation unit 214 can automatically perform a quality evaluation of the generated responses and attempt to regenerate them if they do not meet certain standards.

[0036] The reward granting unit 215 can grant a reward to the template provider when the template is used. For example, the reward granting unit 215 can identify the template provider corresponding to the information identifying the template received by the prompt receiving unit 213 from the template storage unit 231, and process a reward to the identified provider. The reward can be anything that is valuable to the user. The reward may be, for example, points or virtual currency, or it may be crypto assets, digital assets, coupons, etc.

[0037] The reward granting unit 215 may grant rewards by any of the following methods, or a combination thereof. However, it is not limited to these methods.

[0038] (1) Point accrual: A method of awarding points based on the number of times a template is used, and providing rewards based on the accumulated points. (2) Monetary rewards: A method of providing monetary rewards, such as cash or electronic money, based on the number of times a template is used or its evaluation. (3) Service usage rights: A method of providing free or discounted usage rights for paid services within the platform. (4) Ranking display: A method of displaying templates that have made a significant contribution in a ranking format and rewarding the visualized evaluation. (5) Customization privileges: A method of granting platform customization privileges according to the reward level. (6) NFT (Non-Fungible Token): A method of issuing and granting NFTs using blockchain technology, especially for templates of high value. (7) Tiered reward system: A method that introduces a system in which the reward level increases according to the number of times a template is used and its evaluation, allowing users to earn higher rewards.

[0039] If the reward granting unit 215 has stored an original user ID corresponding to the template, it may also grant a reward to the provider who provided the original template before editing (hereinafter referred to as the original provider), as indicated by the original user ID. In this case, the reward granting unit 215 may grant the reward to the provider of the edited template (editor) and the reward to the original provider independently, or it may distribute a portion of the reward to the provider of the edited template (editor) to the original provider.

[0040] The reward distribution unit 215 may implement a multi-layered reward system that takes into account multiple factors such as the number of times a template is used, user ratings, and the number of derived templates.

[0041] Furthermore, the reward distribution unit 215 may be configured to automatically adjust rewards. In this case, the reward distribution unit 215 can automatically adjust the reward rate according to the overall system usage and economic conditions.

[0042] Furthermore, the reward granting unit 215 may set an expiration date for the granted rewards to encourage regular use.

[0043] The template reception unit 216 accepts the submission of templates. The template reception unit 216 can receive the text data of the template from the user terminal 1 of the template creator.

[0044] The template registration unit 217 registers the template received from the creator in the template storage unit 231. The template registration unit 217 searches the template storage unit 231 for templates similar to the received template, and if no similar templates are found, it can register the received template in the template storage unit 231.

[0045] The template editing unit 218 can accept editing requests from editors for templates uploaded by creators. For example, the template selection unit 211 accepts template selections from editors, the template provision unit 212 provides the selected templates to the editor's user terminal 1 so that they can be edited, and the template editing unit 218 can receive the edited templates from the user terminal 1.

[0046] <Operation> Figure 4 is a diagram illustrating the operation of the management server 2.

[0047] The management server 2 accepts the provision of a template (S301), registers the template in association with the provider (S302). If the template has been edited, the template can be registered in association with the original provider. The management server 2 presents a list of templates (S303), accepts the selection of a template (S304), provides the template (S305), accepts a prompt created based on the template (S306), provides the prompt to the large-scale language model to generate an answer (S307), provides the answer to the user (S308), identifies the provider of the used template from the template storage unit 231, and processes a reward for the provider (S309).

[0048] As described above, the information processing system of this embodiment makes it possible to streamline the creation of prompts using templates and to reward providers of prompts that use templates. Therefore, it becomes possible to incentivize the provision of more user-friendly prompts, and consequently, to make the creation of prompts by users more efficient.

[0049] Although these embodiments have been described above, they are intended to facilitate understanding of the present invention and are not intended to limit its interpretation. The present invention can be modified and improved without departing from its spirit, and equivalents thereof are also included.

[0050] For example, the processing performed by each functional unit of the management server 2 described above may be executed by any of the functional units. Furthermore, different functional units may be added to perform some of the processing performed by each of the functional units described above. Also, the functional units of the management server 2 may be distributed across multiple computers.

[0051] Furthermore, the information stored in each memory unit of the management server 2 may be stored in any of the memory units. That is, the information stored in the multiple memory units mentioned above may be stored in a single memory unit, or a portion of the information stored in one memory unit may be stored in another memory unit.

[0052] <Example 1>

[0053] As one modification of this embodiment, the template selection unit 211 can display a list of templates along with information such as the popularity and rating of each template.

[0054] Specifically, the template storage unit 231 can count the number of times each template has been selected and store this information in association with the template. The template selection unit 211 can then retrieve the selection count for each template from the template storage unit 231 and calculate its popularity based on the selection count. The template selection unit 211 can then display the calculated popularity along with a list of templates. The popularity may be, for example, the selection count itself, the selection count over a predetermined period, or the ratio of a particular template's selection count to the total selection count of all templates.

[0055] Furthermore, the template storage unit 231 can store, for each template, the user's evaluation of the response generated from the prompt generated using that template, associating it with the template. The template selection unit 211 can then retrieve the evaluation of each template from the template storage unit 231 and display the retrieved evaluation along with a list of templates. The evaluation may be, for example, the average of the 5-point scale evaluations given by the user to the response, or it may be the number of times the user explicitly evaluated the generated response as excellent.

[0056] <Modification 2>

[0057] As one modification of this embodiment, the template providing unit 212 can add explanations and examples to each part of the template.

[0058] Specifically, the template storage unit 231 can store descriptions and examples of each part (each field) of each template, associating them with the template. The template provisioning unit 212 can then retrieve the template along with its description and example from the template storage unit 231. The template provisioning unit 212 can display each field of the template along with its corresponding description and example. The template provisioning unit 212 can identify each field in the template by enclosing it in special symbols, such as "{}". The template provisioning unit 212 can then recognize each field and use the recognized field name as a key to search for the corresponding description and example from the template storage unit 231. The template provisioning unit 212 can display each field of the template, along with its corresponding description and example, as an edit box or text area.

[0059] Furthermore, the template provider 212 may display the explanation next to the field and the example below the field, or it may display the explanation and example as a tooltip when the mouse cursor hovers over the field.

[0060] <Variation 3>

[0061] As one modification of this embodiment, the management server 2 can enable users to evaluate the answers generated by the answer generation unit 214, and reflect the evaluation results in the distribution of rewards.

[0062] Specifically, the management server 2 may include a response evaluation unit. The response evaluation unit can send the responses generated by the response generation unit 214 to the user terminal 1 along with a user interface for evaluation input. The user interface for evaluation input may, for example, include multiple star-shaped icons for evaluating the quality of the response on a five-point scale, or it may include a "like" button and a comment input field. The response evaluation unit can receive evaluations made by the user.

[0063] Furthermore, the reward granting unit 215 can determine the reward for the template provider based on the evaluation results received by the response evaluation unit. For example, the reward granting unit 215 may grant a reward equal to the basic reward multiplied by a predetermined coefficient if the average value of the evaluations is above a predetermined threshold, or if the number of "likes" is above a predetermined threshold. Conversely, the reward for the template provider may be reduced if the average value of the evaluations is below a predetermined threshold, or if the number of "low ratings" is above a predetermined threshold.

[0064] The reward awarding unit 215 may store the evaluation results in the template storage unit 231 and use them to determine rewards in the future. For example, when determining a reward for a template provider, the reward awarding unit 215 may take into account not only the current evaluation results but also past evaluation results for that template.

[0065] <Modification 4>

[0066] As one modification of this embodiment, the management server 2 may include a template integration unit in addition to the template editing unit 218. The template integration unit may provide a function for integrating similar templates.

[0067] Specifically, the template integration unit can search the template storage unit 231 for templates that are similar to a predetermined degree or higher. The similarity of templates can be calculated, for example, based on the cosine similarity or edit distance of the template text. The template integration unit can send groups of similar templates to the user terminal 1 of the provider that provided those templates and propose template integration to the provider.

[0068] The template integration unit can create an integration proposal by mapping the fields of similar templates and present it to the provider. The template integration unit can accept approval of the integration proposal from the provider. Furthermore, the template integration unit can also accept revisions to the integration proposal from the provider.

[0069] The template integration unit can integrate similar templates into a single template based on the integration proposal approved by the provider. At this time, the template integration unit can record the providers of the integrated template as all the providers of the templates before integration in the template storage unit 231.

[0070] If an integrated template is used, the reward distribution unit 215 can distribute rewards to all providers of the integrated template.

[0071] <Modification 5>

[0072] As one variation of this embodiment, the reward granting unit 215 can link the granting of rewards not only to the number of times the template is used, but also to the quality evaluation of the generated responses.

[0073] Specifically, the response evaluation unit can send the response generated by the response generation unit 214 to the user terminal 1 and receive a quality evaluation of the response from the user terminal 1. The quality evaluation can be, for example, a 5-point rating, the number of "likes," or comments.

[0074] The reward granting unit 215 can determine the reward for the template provider based on the number of times the template is used and the quality evaluation of the responses generated using that template. For example, the reward granting unit 215 can determine a base reward amount based on the number of times the template is used, and then determine the final reward amount by multiplying this base reward amount by a coefficient based on the quality evaluation of the responses.

[0075] The coefficient based on quality evaluation can, for example, be set to a predetermined value if the average evaluation value of the responses is above a predetermined threshold, and to 1 if it is below the threshold. Alternatively, it can be a value proportional to the average evaluation value of the responses. Furthermore, the coefficient based on quality evaluation can be set to a predetermined value if the number of "likes" is above a predetermined threshold, and to 1 if it is below the threshold.

[0076] Furthermore, the reward granting unit 215 can record the history of quality evaluations of responses generated using templates in the template storage unit 231. The reward granting unit 215 can also use the history of quality evaluations to determine rewards for template providers. For example, the reward granting unit 215 can determine a coefficient based on quality evaluations using the average value of quality evaluations over a predetermined period in the past.

[0077] <Disclosure Items> Furthermore, this disclosure also includes the following configurations. [Item 1] A template storage unit that stores prompt templates for a large-scale language model in association with the template provider, A template selection unit that accepts the selection of one template from the aforementioned templates, A template provisioning unit that provides the selected template to the user, A prompt receiving unit that receives a prompt created from the user based on the template, A response generation unit provides the received prompt to the large-scale language model to generate a response, A reward granting unit identifies the provider of the template used to create the received prompt from the template storage unit and grants a reward to the identified provider, An information processing system characterized by comprising the following features. [Item 2] The information processing system described in item 1, The prompt receiving unit receives the prompt along with information that identifies the template. An information processing system characterized by the following. [Item 3] The information processing system described in item 1, The reward granting unit searches the template storage unit for a template similar to the received prompt. An information processing system characterized by the following. [Item 4] The information processing system described in item 1, A template reception unit that accepts the aforementioned templates, A template registration unit searches the template storage unit for templates similar to the received template, and if no similar template is found, registers the received template in the template storage unit. An information processing system characterized by comprising the following features. [Item 5] The steps include storing a template for a prompt to be given to a large-scale language model in a template storage unit, associating it with the template provider, A step that accepts the selection of one template from the above templates, The steps include providing the selected template to the user, A step of receiving a prompt from the user based on the template, The steps include providing the received prompt to the large-scale language model to generate a response, The steps include: identifying the provider of the template used to create the received prompt from the template storage unit, and rewarding the identified provider; An information processing method characterized by a computer executing the following. [Item 6] The steps include storing a template for a prompt to be given to a large-scale language model in a template storage unit, associating it with the template provider, A step that accepts the selection of one template from the above templates, The steps include providing the selected template to the user, A step of receiving a prompt from the user based on the template, The steps include providing the received prompt to the large-scale language model to generate a response, The steps include: identifying the provider of the template used to create the received prompt from the template storage unit, and rewarding the identified provider; A program that causes a computer to execute something. [Explanation of symbols]

[0078] 1 User terminal 2 Management Server

Claims

1. A template storage unit that stores prompt templates for a large-scale language model in association with the template provider, A template selection unit that accepts the selection of one template from the aforementioned templates, A template provisioning unit that provides the selected template to the user, A prompt receiving unit that receives a prompt created from the user based on the template, A response generation unit provides the received prompt to the large-scale language model to generate a response, A reward granting unit identifies the provider of the template used to create the received prompt from the template storage unit and grants a reward to the identified provider, An information processing system characterized by comprising the following features.

2. The information processing system according to claim 1, The prompt receiving unit receives the prompt along with information that identifies the template. An information processing system characterized by the following.

3. The information processing system according to claim 1, The reward granting unit searches the template storage unit for a template similar to the received prompt. An information processing system characterized by the following.

4. The information processing system according to claim 1, A template reception unit that accepts the aforementioned templates, A template registration unit searches the template storage unit for templates similar to the received template, and if no similar template is found, registers the received template in the template storage unit. An information processing system characterized by comprising the following features.

5. The steps include storing a template for a prompt to be given to a large-scale language model in a template storage unit, associating it with the template provider, A step that accepts one selection from the aforementioned templates, The steps include providing the selected template to the user, A step of receiving a prompt from the user based on the template, The steps include providing the received prompt to the large-scale language model to generate a response, The steps include: identifying the provider of the template used to create the received prompt from the template storage unit, and rewarding the identified provider; An information processing method characterized by a computer executing the following.

6. The steps include storing a template for a prompt to be given to a large-scale language model in a template storage unit, associating it with the template provider, A step that accepts one selection from the aforementioned templates, The steps include providing the selected template to the user, A step of receiving a prompt from the user based on the template, The steps include providing the received prompt to the large-scale language model to generate a response, The steps include: identifying the provider of the template used to create the received prompt from the template storage unit, and rewarding the identified provider; A program that causes a computer to execute something.

Images