Information processing device, information processing method, and information processing program

Abstract

To provide an information processing device, an information processing method, and an information processing program that can generate highly accurate responses to user requests. [Solution] The information processing device according to the present invention comprises a reception unit that receives user requests, a selection unit that selects an agent corresponding to the request from among agents that generate responses to the requests, and a generation unit that generates content to be provided to the user based on the response of the selected agent to the request.

Description

Technical Field

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

Background Art

[0002] In recent years, technology in which an agent configured with artificial intelligence (AI: Artificial Intelligence) receives a request from a user and provides content serving as a response to the request has been becoming widespread (see, for example, Patent Document 1).

Prior Art Literature

Patent Literature

[0003]

Patent Document 1

Summary of the Invention

Problem to be Solved by the Invention

[0004] However, in the conventional technology, there is room for improvement in terms of generating a highly accurate response to a user's request.

[0005] The present application has been made in view of the above circumstances, and an object of the present application is to provide an information processing apparatus, an information processing method, and an information processing program that can generate a highly accurate response to a user's request.

Means for Solving the Problem

[0006] The information processing device according to the present invention comprises a receiving unit that receives user requests, a selection unit that selects an agent corresponding to the request from among agents that generate responses to the requests, and a generation unit that generates content to be provided to the user based on the response of the selected agent to the request. Note that some or all of the above-described solutions may be implemented using various artificial intelligence technologies and methods (including, but not limited to, the use of various generative AI models realized using machine learning, deep learning, large-scale language models, etc.). [Effects of the Invention]

[0007] According to one embodiment, the system has the effect of generating highly accurate responses to user requests. [Brief explanation of the drawing]

[0008] [Figure 1] Figure 1 shows an example of the operation of an information processing system including an information processing device according to the embodiment. [Figure 2] Figure 2 is a block diagram showing an example configuration of an information processing system according to the embodiment. [Figure 3] Figure 3 shows an example of the configuration of an information processing device according to the embodiment. [Figure 4] Figure 4 shows an example of user information. [Figure 5] Figure 5 shows an example of model information. [Figure 6] Figure 6 shows an example of an instruction manual database. [Figure 7] Figure 7 is a flowchart showing the processing procedure of the information processing device executed by the information processing device according to the embodiment. [Figure 8] Figure 8 shows an example of a hardware configuration. [Modes for carrying out the invention]

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

[0010] (Embodiment) First, an overview of the processing performed by the information processing device according to the embodiment will be described using Figure 1. Figure 1 is a diagram showing the processing performed by the information processing device according to the embodiment. Note that Figure 1 shows an example of the operation of the information processing system S including the information processing device 1 according to the embodiment. The information processing device 1 also stores an AI model that generates answers to user requests. In this disclosure, the AI ​​model is defined as Main AGA1, First Sub AGA2, Second Sub AGA3, and Video Generation AIA4. Note that the AI ​​model is not limited to one generated by the information processing device 1 through learning or set (rule-based) by the business operator managing the information processing device 1; an AI model stored on another server outside the information processing device 1 may also be used.

[0011] Furthermore, while this disclosure uses the example of a user's request relating to a travel plan, the user's request is not limited to a travel plan and may be any request at all.

[0012] As shown in Figure 1, the information processing system S according to this embodiment includes an information processing device 1 and a user terminal 100. The user terminal 100 is a terminal device owned by a user who makes requests (such as questions) to the information processing device 1.

[0013] In this disclosure, the information processing device 1 has three main characteristic functions.

[0014] First, the information processing apparatus 1 has a function of receiving a user's request, selecting a response agent corresponding to the request from among response agents that generate a response to the request, and generating content to be provided to the user based on the response to the request of the selected response agent.

[0015] Second, the information processing apparatus 1 has a function of extracting an instruction sheet corresponding to a command to an agent from a database including instruction sheets indicating instruction contents, inputting a prompt including the extracted instruction sheet to the agent, and thereby generating the agent's response to the command.

[0016] Third, the information processing apparatus 1 has a function of identifying a proposal target corresponding to a user's request based on the user's request, selecting a provision image to be provided from among target images corresponding to the identified proposal target, and providing a video generated based on the selected provision image to the user.

[0017] Hereinafter, a detailed description will be given focusing on the above three functions with reference to Figure 1.

[0018] First, prior to describing the overall processing of the information processing system S, the AI models stored in the information processing apparatus 1 (main AGA1, first sub-AGA2, second sub-AGA3, and video generation AIA4) will be described.

[0019] The main AGA1 is, for example, an AI agent implemented by an LLM (Large Language Model), and is an agent that receives requests from users and provides responses to the requests. The main AGA1 also issues response generation instructions to the first sub-AGA2, the second sub-AGA3, and the video generation AIA4, and generates content to be provided to the user based on the responses generated by the first sub-AGA2, the second sub-AGA3, and the video generation AIA4.

[0020] The first sub-AGA2 is an AI agent implemented by, for example, LLM, which operates according to prompts input from the main AGA1 and generates responses corresponding to those prompts. The first sub-AGA2 exists for each type of response it outputs. For example, the first sub-AGA2 may exist for each field of response it is responsible for, such as a model that outputs responses about travel plans, a model that outputs responses about shopping, or a model that outputs responses about work support.

[0021] The second sub-AGA3 is an AI agent implemented, for example, by a rule-based AI model. It operates according to prompts input from the main AGA1, generating responses that correspond to those prompts. Specifically, the second sub-AGA3 outputs responses that follow rules predetermined by the designer. For example, the second sub-AGA3 could be a model that outputs search results for accommodations or a model that outputs travel routes to a destination, outputting responses that conform to the rules.

[0022] The video generation AIA4 is a generation AI implemented, for example, by a multimodal generation AI. It takes video generation prompts and images input from the main AGA1 as input, and uses those images to generate a video corresponding to the prompt.

[0023] The information processing device 1 disclosed herein performs various processes described below using the AI ​​model described above.

[0024] Specifically, first, the information processing device 1 receives a request from the user (step S1). The request may be, for example, a question about travel or a request to create a travel plan. The information processing device 1 receives the request by interacting with the user via the main AGA1. The interaction may be, for example, voice interaction or text interaction (chat, etc.). In addition, the user can also provide images related to the request to the main AGA1 via the user terminal 100 during the interaction.

[0025] Next, the information processing device 1 analyzes the received request using the main AGA1 and classifies it into tasks (step S2). Specifically, the main AGA1 of the information processing device 1 repeatedly engages in dialogue with the user to delve deeper into the request, thereby breaking it down (classifying) into multiple tasks. More specifically, the information processing device 1 classifies the request into tasks that can be answered by the first sub-AGA2 and tasks that can be answered by the second sub-AGA3. The method of delving deeper into the request is, for example, by explicitly giving instructions (guidelines) to the main AGA1.

[0026] Main AGA1, for example, if the user requests "I want to go on a relaxing trip," will generate patterns related to travel options (such as a seaside resort or a quiet mountain area) and present them to the user through dialogue to obtain a response. In other words, Main AGA1 performs linguistic analysis of the request, and if the request is highly abstract, it will repeat the dialogue until the request becomes more specific by asking questions to delve deeper into the request (in this case, since only the requests for "relaxation" and "travel" are known, it will delve deeper into "relaxation" and "travel" to make them more specific). Then, if Main AGA1 receives a response such as a seaside resort, it will obtain information such as the timing and duration of the trip and the desired destination through dialogue. In this case, Main AGA1 will generate a prompt containing information such as the desired timing and duration and destination, and input it into the first sub-AGA2. In other words, Main AGA1 classifies tasks that cannot be answered by the second sub-AGA3 as tasks for the first sub-AGA2. Furthermore, if Main AGA1 receives requests from users for recommended accommodations after proposing a plan, these requests will be classified as a task under Sub-AGA3.

[0027] Next, the information processing device 1 selects a sub-AG corresponding to each classified task (step S3). Specifically, the main AGA1 selects the first sub-AGA2 for tasks classified as tasks of the first sub-AGA2, and selects the second sub-AGA3 for tasks classified as tasks of the second sub-AGA3.

[0028] Next, the information processing device 1 generates prompts to be input to each selected sub-AG based on each task (step S4). Specifically, the main AGA1 generates a prompt for the first sub-AGA2 to generate an answer to the task corresponding to the first sub-AGA2. The main AGA1 also generates a prompt for the second sub-AGA3 to generate an answer to the task corresponding to the second sub-AGA3.

[0029] Next, the information processing device 1 takes prompts as input to each sub-AG and generates answers for each task (step S5). Specifically, the first sub-AGA2 takes the prompt generated by the main AGA1 as input and outputs (generates) answers to the tasks. The second sub-AGA3 also takes the prompt generated by the main AGA1 as input and outputs (generates) answers to the tasks.

[0030] Next, the information processing device 1 generates content to be provided to the user based on the answers of each sub-AG obtained in step S5 (step S6). For example, if a travel plan is generated as the answer to the first sub-AGA2, the main AGA1 generates text content indicating the travel plan as the response in the dialogue. Also, if a recommended accommodation is generated as the answer to the second sub-AGA3, the main AGA1 generates text content indicating the accommodation as the response in the dialogue. Note that the main AGA1 may generate content for each of the answers to the first sub-AGA2 and the second sub-AGA3, or it may generate one piece of content based on each of the answers.

[0031] Furthermore, Main AGA1 may, for example, input a response generated by Second Sub AGA3 as a prompt into First Sub AGA2, and generate content based on the response output from First Sub AGA2. For example, Main AGA1 collects user evaluations (reviews, etc.) related to the needs of the proposed item (e.g., a hotel) using Second Sub AGA3, and generates a summary of the user evaluations collected by Second Sub AGA3 as the response for First Sub AGA2. Then, based on the summary of user evaluations which is the response from First Sub AGA2, Main AGA1 may, for example, determine whether or not to propose the proposed item to the user, or generate content by changing the summary to an expression that matches the user's needs.

[0032] Next, the information processing device 1 extracts the images to be provided from among the target images corresponding to the proposed target identified from the user's request received in step S1 (step S7). Specifically, first, main AGA1 identifies the proposed target from the user's request. The proposed target is, for example, accommodation. In addition to identifying the proposed target from the user's request, it may also be identified from the response content generated in step S5 above or the content generated in step S6 above. Note that the user's request received in step S7 may be different from the user's request received in step S1. In other words, although step S7 is described here as a series of processes following steps S1 to S6, it is not limited to this, and new user requests may be received as a series of processes starting from this step. Next, main AGA1 extracts the target images corresponding to the identified proposed target. For example, main AGA1 extracts target images of accommodation from images posted on the accommodation's reservation site or homepage. Next, Main AGA1 extracts the images to be provided to the user from the extracted target images. For example, Main AGA1 extracts images that match the user's request based on the metadata attached to the target images. Metadata includes, for example, alt (angle information), type (type of image content: food, guest room, lobby, etc.), and description (detailed image information: 10th floor restaurant "AA store," etc.).

[0033] Specifically, MainAGA1 determines whether the metadata matches the user's requirements and selects a target image with matching metadata as the provided image. Furthermore, if the target image lacks metadata, MainAGA1 lowers its selection priority. Specifically, the more missing metadata items MainAGA1 has, the lower its selection priority. Additionally, if the image's aspect ratio is included as metadata, MainAGA1 may select the provided image based on that aspect ratio. For example, MainAGA1 will not select an image if its aspect ratio is not a predetermined standard ratio (e.g., 4:3 or 16:9). Note that the standard ratio can be defined by numerical values, or it can be set to specify the image shape without limiting the numerical ratio, such as "portrait" or "landscape."

[0034] Furthermore, MainAGA1 may, for example, use a multimodal generation AI to extract images that match the user's requests based on the image analysis results. For example, MainAGA1 may accept images and text (e.g., "I want a hotel room with this kind of atmosphere") as user requests, input the accepted images and text along with target images into the multimodal generation AI, calculate a score indicating the relationship between the requests and target images, and select images based on that score. Note that user requests are not limited to accepting both images and text; it is also possible to accept only images or only text as user requests.

[0035] Furthermore, MainAGA1 may self-evaluate whether the selected provided image matches the user's request. For example, MainAGA1 may calculate a score indicating the relevance between the metadata attached to the provided image and the user's request, and confirm it as a provided image if the score is above a threshold. Alternatively, MainAGA1 may, without using metadata, convert the provided image into text using, for example, LLM, calculate a score indicating the relevance between the converted text and the request, and confirm it as a provided image if the score is above a threshold.

[0036] Next, the information processing device 1 generates a prompt to be input to the video generation AIA4 (step S8). Specifically, the main AGA1 determines the video generation policy based on the user's request and generates a prompt to instruct video generation according to the determined generation policy. The generation policy is a policy that indicates the purpose for which the video will be used, such as "attractive video" or "explanatory video". The main AGA1 generates prompts using templates or prompt generation LLM (not shown). Specifically, the main AGA1 selects a template corresponding to the generation policy from among several pre-prepared prompt templates and generates that template as a prompt. In other words, the main AGA1 extracts an instruction document (template) corresponding to the command to the agent from an instruction document database (DB) containing instruction documents (templates) that indicate the instruction content (generation policy), and generates a prompt that includes the content described in the extracted instruction document (template). Furthermore, the main AGA1 inputs the prompts from the prompt generation LLM along with the provided image into the prompt generation LLM, and generates prompts for the video generation AI from the output results of the prompt generation LLM. In this disclosure, the information processing device 1 has shown examples of prompt generation using templates and prompt generation using a prompt generation LLM (not shown) for video generation, but for example, prompt generation for travel plans, i.e., prompts to be input to the first sub-AGA2 and second sub-AGA3, may also be performed using templates or prompt generation using a prompt generation LLM (not shown).

[0037] Furthermore, the instruction database (or the instruction documents it contains) may include instructions for, for example, to be added to prompts. The description of such instructions may be divided into two parts: an "instruction judgment part" and an "instruction content part." The "instruction judgment part" includes text for determining when to use (select) the instruction content described in the "instruction content part." Examples of such text may include, but are not limited to, descriptions such as, "When you are asked to plan a trip based on vague requests. For example, 'I want to relax,' 'I want to see a sea of ​​clouds,' 'I want to refresh my mind,' etc.," or "When you have clear requests for specific accommodations and it is necessary to strictly adhere to those conditions." It may also include text related to multiple "instruction judgment parts," in which case there would be two instruction documents in the instruction database. In other words, there may be multiple instruction documents in the instruction database that are set together with an "instruction judgment part" and an "instruction content part."

[0038] The information processing device 1, based on the user's request (prompt instruction), scores the degree of matching with all "instruction judgment parts" in the instruction document DB using an LLM (it may use the LLM used previously or a different LLM), and uses the instruction document with the highest score, adding the content described in its "instruction content" to the prompt that is finally generated for processing.

[0039] Next, the information processing device 1 generates a video related to the proposed subject using the generated prompt and the provided image as input (step S9). Specifically, when the video generation AIA4 receives the prompt generated by the main AGA1 and the provided image as input, it generates a video related to the proposed subject using the provided image.

[0040] Next, the information processing device 1 provides the content generated in step S6 and the video generated in step S9 to the user terminal 100 (step S10). Specifically, the main AGA1 provides the generated content and video as a response to requests received from the user during interaction with the user.

[0041] Next, an example of the configuration of the information processing system S according to the embodiment will be described using Figure 2. Figure 2 is a block diagram showing an example of the configuration of the information processing system S according to the embodiment. As shown in Figure 2, in the information processing system S according to the embodiment, the information processing device 1 and a plurality of user terminals 100 are connected to a network N by wire or wireless. The network N is, for example, a network such as the Internet, a WAN (Wide Area Network), or a LAN (Local Area Network).

[0042] The information processing device 1 is a server device that executes the information processing method according to the embodiment. The information processing device 1 has the function of receiving a user request, selecting a response agent that corresponds to the request from among response agents that generate responses to the request, and generating content to be provided to the user based on the response of the selected response agent to the request.

[0043] Furthermore, the information processing device 1 has the function of extracting instruction documents corresponding to commands to the agent from a database containing instruction documents that indicate the content of the instructions, and inputting a prompt containing the extracted instruction document to the agent, thereby generating the agent's response to the command.

[0044] Furthermore, the information processing device 1 has the function of identifying a proposed target that corresponds to the user's request based on the user's request, selecting a provided image from among the target images corresponding to the identified proposed target, and providing the user with a video generated based on the selected provided image.

[0045] Furthermore, the information processing device 1 is an information processing device that interacts with multiple user terminals 100 and provides API (Application Programming Interface) services for various applications (hereinafter referred to as "apps") such as chat services, as well as various data to the multiple user terminals 100, and is realized by a server device or a cloud system.

[0046] Furthermore, the information processing device 1 may be an information processing device that provides some kind of web service online to multiple user terminals 100. For example, the information processing device 1 may provide services such as internet connection, search services, SNS (Social Networking Service), e-commerce (EC), electronic payment, online games, online banking, online trading, accommodation / ticket reservations, video / music distribution, news, maps, route search, route guidance, route information, service information, and weather forecasts as web services. In practice, the information processing device 1 may cooperate with various servers that provide the above-mentioned web services and act as an intermediary for web services, or it may be responsible for processing web services.

[0047] The user terminal 100 is a terminal device owned by the user. The user terminal 100 can be any type of terminal device, such as a smartphone, desktop PC, notebook PC, or tablet PC. The user terminal 100 transmits various types of information to the information processing device 1, etc., and receives information provided by the information processing device 1, etc.

[0048] Next, with reference to Figure 3, an example configuration of the information processing device 1 will be described.

[0049] Figure 3 is a diagram showing an example configuration of an information processing device 1 according to an embodiment. As shown in Figure 3, the information processing device 1 has a communication unit 2, a control unit 3, and a storage unit 4. The control unit 3 includes a reception unit 31, a selection unit 32, an extraction unit 33, a generation unit 34, a specification unit 35, and a provision unit 36. The storage unit 4 stores user information 41, model information 42, and an instruction database (DB) 43.

[0050] The communication unit 2 is implemented, for example, by a NIC (Network Interface Card). The communication unit 2 is connected to the network by wire or wireless connection.

[0051] The control unit 3 is a controller and is implemented by a processor such as a CPU (Central Processing Unit) or MPU (Micro Processing Unit) executing various programs (corresponding to an example of an information processing program) stored in the memory device inside the information processing device 1, using RAM or the like as a working area. Alternatively, the control unit 3 is a controller and may be implemented by an integrated circuit such as an ASIC (Application Specific Integrated Circuit), FPGA (Field Programmable Gate Array), or GPGPU (General Purpose Graphic Processing Unit).

[0052] The memory unit 4 is implemented by, for example, semiconductor memory elements such as RAM (Random Access Memory) and flash memory, or by storage devices such as hard disks and optical discs.

[0053] User information 41 is information about the user.

[0054] Figure 4 shows an example of user information 41. As shown in Figure 4, user information 41 includes items such as "User ID," "Attribute Information," and "Advanced History."

[0055] "User ID" is identification information that identifies a user. "Attribute Information" is information about the user's attributes. Attribute information includes, for example, psychographic attributes and demographic attributes. "Behavioral History" is information about the user's behavioral history, such as search behavior, purchase behavior, and interaction behavior with Main AGA1.

[0056] Model information 42 is information about the AI ​​model.

[0057] Figure 5 shows an example of model information 42. As shown in Figure 5, model information 42 includes items such as "Model ID," "Model Name," and "Model Type."

[0058] "Model ID" is identification information that identifies the AI ​​model. "Model Name" is the name of the AI ​​model. "Model Type" is information that indicates the type of AI model.

[0059] Instruction DB43 is a database of instructions regarding the rules for generating prompts to be input to the video generation AI.

[0060] Figure 6 shows an example of the instruction sheet DB43. As shown in Figure 6, the instruction sheet DB43 includes items such as "instruction sheet ID," "number of sheets rule," "arrangement rule," and "effect rule."

[0061] "Instruction ID" is identification information that identifies the instruction sheet. "Number of Images Rule" is information that indicates the conditions (rules) for determining the number of provided images to be used in the generated video. "Arrangement Rule" is information that indicates the conditions (rules) for determining the arrangement of provided images in the video (the order in which they are displayed in the video). "Effect Rule" is information that indicates the conditions (rules) for determining the effects (wipe, etc.) to be added to the provided images in the generated video.

[0062] Next, we will explain the functions of the control unit 3 (reception unit 31, selection unit 32, extraction unit 33, generation unit 34, identification unit 35, and provision unit 36).

[0063] The control unit 3 executes the functions of the reception unit 31, selection unit 32, extraction unit 33, generation unit 34, identification unit 35, and provision unit 36, thereby executing the processes of the information processing device 1 as described in Figure 1.

[0064] The reception unit 31 performs the reception processing of various types of information as described in Figure 1. The selection unit 32 performs the selection processing of various types of information as described in Figure 1. The extraction unit 33 performs the extraction processing of various types of information as described in Figure 1. The generation unit 34 performs the generation processing of various types of information as described in Figure 1. The identification unit 35 performs the identification processing of various types of information as described in Figure 1. The provision unit 36 ​​performs the provision processing of various types of information as described in Figure 1.

[0065] Next, the processing procedure of the process executed by the information processing device 1 according to the embodiment will be described using Figure 7. Figure 7 is a flowchart showing the processing procedure of the process executed by the information processing device 1 according to the embodiment.

[0066] As shown in Figure 7, the control unit 3 first receives the user's request using the main AGA1 (step S101).

[0067] Next, the control unit 3 uses the main AGA1 to classify the received requests into tasks (step S102).

[0068] Next, the control unit 3 uses the main AGA1 to select a subAG for each classified task (step S103).

[0069] Next, the control unit 3 generates a prompt for each sub-AG using the main AGA1 (step S104).

[0070] Next, the control unit 3 generates an answer to the task using the sub-AG (step S105).

[0071] Next, the control unit 3 uses the main AGA1 to generate content to be provided to the user based on the answers from the subAG (step S106).

[0072] Next, the control unit 3 uses the main AGA1 to identify the target of the proposal based on the user's request (step S107).

[0073] Next, the control unit 3 uses the main AGA1 to select the image to be provided from the target image of the proposal (step S108).

[0074] Next, the control unit 3 uses the video generation AIA4 to generate a video related to the proposed object using the provided image and prompt as input (step S109).

[0075] Next, the control unit 3 provides the generated content and the proposed video to the user (step S110), and then terminates the process.

[0076] 〔others〕 Furthermore, some of the processes described as being performed automatically in the above embodiments can be performed manually. Alternatively, all or part of the processes described as being performed manually can be performed automatically by known methods. In addition, the processing procedures, specific names, and various data and parameters shown in the above documents and drawings can be changed at will unless otherwise specified. For example, the various information shown in each figure is not limited to the information shown.

[0077] Furthermore, the components of each illustrated device are functionally conceptual and do not necessarily need to be physically configured as shown. In other words, the specific forms of distribution and integration of each device are not limited to those shown, and all or part of them can be functionally or physically distributed and integrated in any unit according to various loads and usage conditions.

[0078] For example, some or all of the storage unit 4 shown in Figure 3 may be stored in a storage server or the like, rather than being held by each device. In this case, each device obtains various information by accessing the storage server.

[0079] [Hardware configuration] Furthermore, the information processing device 1 according to the above embodiment is realized by a computer 1000 having a configuration such as that shown in Figure 8. Figure 8 is a diagram showing an example of a hardware configuration. The computer 1000 is connected to an output device 1010 and an input device 1020, and has a configuration in which an arithmetic unit 1030, a primary storage device 1040, a secondary storage device 1050, an output IF (Interface) 1060, an input IF 1070, and a network IF 1080 are connected by a bus 1090.

[0080] The arithmetic unit 1030 operates based on programs stored in the primary storage device 1040 and the secondary storage device 1050, as well as programs read from the input device 1020, and executes various processes. The primary storage device 1040 is a memory device, such as RAM, that temporarily stores data used by the arithmetic unit 1030 for various calculations. The secondary storage device 1050 is a storage device where data used by the arithmetic unit 1030 for various calculations and various databases are registered, and is implemented using ROM (Read Only Memory), HDD (Hard Disk Drive), flash memory, etc.

[0081] Output IF1060 is an interface for transmitting information to be output to output devices 1010, which output various types of information such as monitors and printers. It is implemented using connectors of standards such as USB (Universal Serial Bus), DVI (Digital Visual Interface), and HDMI (High Definition Multimedia Interface). Input IF1070 is an interface for receiving information from various input devices 1020, such as mice, keyboards, and scanners. It is implemented using, for example, USB.

[0082] The input device 1020 may also be a device that reads information from, for example, an optical recording medium such as a CD (Compact Disc), DVD (Digital Versatile Disc), or PD (Phase Change Rewritable Disk), a magneto-optical recording medium such as an MO (Magneto-Optical disk), tape media, magnetic recording media, or semiconductor memory. Furthermore, the input device 1020 may also be an external storage medium such as a USB memory stick.

[0083] Network IF1080 receives data from other devices via network N and sends it to the arithmetic unit 1030, and also transmits data generated by the arithmetic unit 1030 to other devices via network N.

[0084] The arithmetic unit 1030 controls the output device 1010 and the input device 1020 via the output IF 1060 and the input IF 1070. For example, the arithmetic unit 1030 loads a program from the input device 1020 or the secondary storage device 1050 onto the primary storage device 1040 and executes the loaded program.

[0085] For example, when computer 1000 functions as information processing device 1, the arithmetic unit 1030 of computer 1000 realizes the functions of control unit 3 by executing a program loaded on primary storage device 1040.

[0086] 〔effect〕 As described above, the information processing device 1 according to the embodiment includes a reception unit 31 that receives user requests, a selection unit 32 that selects an agent corresponding to the request from among agents that generate responses to the requests, and a generation unit 34 that generates content to be provided to the user based on the response of the selected agent to the request. According to this disclosure, it is possible to generate highly accurate responses to user requests.

[0087] Furthermore, the information processing device 1 according to the embodiment includes an extraction unit 33 that extracts instruction documents corresponding to commands to an agent from an instruction document database containing instruction documents showing the content of the instructions, and a generation unit 34 that generates the agent's response to the command by inputting a prompt containing the extracted instruction document to the agent. According to this disclosure, the accuracy of the agent's response to the command can be improved.

[0088] Furthermore, the information processing device 1 according to this embodiment includes a specification unit 35 that identifies a proposed target corresponding to a user's request based on the user's request, a selection unit 32 that selects a provided image to be provided from among the target images corresponding to the identified proposed target, and a provision unit 36 ​​that provides the user with a video generated based on the selected provided image. According to this disclosure, the accuracy of the agent's response to commands can be improved. High-precision videos can be generated in response to the user's request.

[0089] Although some embodiments of the present invention have been described in detail above with reference to the drawings, these are illustrative examples, and the present invention can be implemented in various other forms with modifications and improvements based on the knowledge of those skilled in the art, starting with the embodiments described in the disclosure section of the invention.

[0090] 〔others〕 Furthermore, among the processes described in the above embodiments, all or part of the processes described as being performed automatically can be performed manually, or all or part of the processes described as being performed manually can be performed automatically by known methods. In addition, the processing procedures, specific names, and various data and parameters shown in the above document and drawings can be changed at will unless otherwise specified. For example, the various information shown in each figure is not limited to the information shown.

[0091] Furthermore, the components of each illustrated device are functionally conceptual and do not necessarily need to be physically configured as shown. In other words, the specific forms of distribution and integration of each device are not limited to those shown, and all or part of them can be functionally or physically distributed and integrated in any unit according to various loads and usage conditions.

[0092] Furthermore, each of the processes described in the embodiments above can be combined as appropriate, provided that the processing content does not contradict each other.

[0093] Furthermore, the terms "section, module, unit" used above can be replaced with "means" or "circuit," etc. For example, control unit 3 can be replaced with control means or control circuit. [Explanation of Symbols]

[0094] 1. Information Processing Device 2 Communications Department 3. Control Unit 4 Storage section 31 Reception Department 32 Selection Section 33 Extraction part 34 Generation part 35 Specific part 36 Providing Department 41 User Information 42 Model Information 100 user terminals S Information Processing System

Claims

1. A reception desk that handles user requests, A selection unit that selects an agent corresponding to the request from among agents that generate answers to the request, A generation unit that generates content to be provided to the user based on the selected agent's response to the request, Equipped with, The aforementioned agent, This includes a main agent that receives the aforementioned request and a sub-agent that generates a response to the aforementioned request. The aforementioned selection unit is Using the main agent, select the sub-agent that corresponds to the request, The generating unit is The main agent generates the content based on the response to the request generated by the sub-agent. Information processing device.

2. The aforementioned selection unit is Using the main agent, the requirements are analyzed and classified into multiple tasks, and a sub-agent corresponding to each of the multiple tasks is selected. The information processing apparatus according to claim 1.

3. The aforementioned subagent is It is an LLM or rule-based AI, The aforementioned selection unit is Using the main agent, select the LLM or the sub-agent of the rule-based AI for each of the multiple tasks. The information processing apparatus according to claim 2.

4. The generating unit is Using the main agent, the response generated by the rule-based AI is input as a prompt to the sub-agent of the LLM, and the content is generated based on the response generated by the LLM. The information processing apparatus according to claim 3.

5. The aforementioned rule-based AI is The results of the user evaluations for the proposed solution are collected and used to generate the above response. The aforementioned LLM is, The summary of the collected user evaluations is generated as the response. The information processing apparatus according to claim 4.

6. The aforementioned main agent is LLM, The aforementioned reception unit is The main agent is used to interact with the user and accept the request. The information processing apparatus according to claim 1.

7. The aforementioned request concerns travel, The generating unit is The aforementioned content generates a travel plan that meets the aforementioned request and the accommodations included in the travel plan. The information processing apparatus according to claim 1.

8. A method of information processing performed by a computer, The process of receiving user requests, A selection step of selecting an agent that corresponds to the request from among agents that generate answers to the request, A generation process that generates content to be provided to the user based on the selected agent's response to the request, Includes, The aforementioned agent, This includes a main agent that receives the aforementioned request and a sub-agent that generates a response to the aforementioned request. The aforementioned selection step is, Using the main agent, select the sub-agent that corresponds to the request, The aforementioned generation step is The main agent generates the content based on the response to the request generated by the sub-agent. Information processing methods.

9. The procedure for receiving user requests, A selection procedure for selecting an agent that corresponds to the request from among agents that generate responses to the request, A generation procedure for generating content to be provided to the user based on the selected agent's response to the request, Have the computer run it, The aforementioned agent, This includes a main agent that receives the aforementioned request and a sub-agent that generates a response to the aforementioned request. The above selection procedure is, Using the main agent, select the sub-agent that corresponds to the request, The aforementioned generation procedure is: The main agent generates the content based on the response to the request generated by the sub-agent. Information processing program.

Images