Information processing device, information processing method, and computer-readable recording medium

Abstract

The present invention provides an evaluation result of a user with respect to another user and / or an evaluation result of the other user with respect to the user without causing the user to find an interaction time.　This information processing device 100 comprises: an agent creation unit 140 that creates AI agents on the basis of agent information; an interaction control unit 160 that causes a first AI agent created on the basis of first agent information and a second AI agent created on the basis of second agent information to interact with each other under a prescribed purpose, and acquires a first evaluation result of the second AI agent with respect to the first AI agent and / or a second evaluation result of the first AI agent with respect to the second AI agent; a result creation unit 170 that creates a matching result on the basis of the first evaluation result and / or the second evaluation result; and output units 120, 180 that output the matching result.

Description

Information Processing Apparatus, Information Processing Method, and Computer-Readable Recording Medium Cross-Reference to Related Applications

[0001] This application is based on Japanese Patent Application No. 2025-9167 filed on January 22, 2025, and Japanese Patent Application No. 2025-60953 filed on April 2, 2025, the contents of which are incorporated herein by reference.

[0002] The present disclosure relates to an information processing apparatus, an information processing method, and a computer-readable recording medium.

[0003] In recent years, services using AI agents that interact with people have been increasing. In the system described in Patent Document 1, a virtual animated interactive agent can be rendered for visual and auditory communication with one or more users within an application. For example, a conversational artificial intelligence (AI) assistant can be rendered and displayed for visual communication in addition to auditory communication with an end user. Thereby, the system aims to utilize the visual area in addition to the auditory area, including the interaction of the AI assistant with the virtual environment in which the AI assistant is rendered, so that the AI assistant can communicate with the user more clearly. In addition, the system aims to utilize voice, video, and / or text input from the user to determine the user's requests, mood, gestures, and / or posture for the AI assistant to respond more accurately to the user and interact with the user.

[0004] Japanese Patent Publication No. 2023-525173

[0005] However, in the prior art including the system described in Patent Document 1, it is premised that a person and an AI assistant interact with each other, and it is impossible to provide the evaluation result of another user with respect to the user or the evaluation result of the user with respect to another user without squeezing out the conversation time for the user.

[0006] Therefore, this disclosure aims to provide users with evaluation results from other users and / or evaluation results from other users of that user, without requiring the user to set aside time for dialogue.

[0007] An information processing device according to one aspect of the present disclosure includes: an agent creation unit that creates an AI agent based on agent information; a dialogue control unit that causes a first AI agent created based on first agent information and a second AI agent created based on second agent information to interact for a predetermined purpose and obtains a first evaluation result of the second AI agent with respect to the first AI agent and / or a second evaluation result of the first AI agent with respect to the second AI agent; a result creation unit that creates a matching result based on the first evaluation result and / or the second evaluation result; and an output unit that outputs the matching result. An information processing device according to one aspect of the present disclosure includes: an agent creation unit that creates an AI agent based on agent information; a dialogue control unit that causes a first AI agent created based on first agent information and a second AI agent created based on second agent information to interact for a predetermined purpose and obtains a first evaluation result for the first AI agent that reflects the consciousness load of the second AI agent and / or a second evaluation result for the second AI agent that reflects the consciousness load of the first AI agent; a result creation unit that creates a matching result based on the first evaluation result and / or the second evaluation result; and an output unit that outputs the matching result.

[0008] According to this disclosure, it becomes possible to provide users with evaluation results from other users and / or evaluation results from other users of that user, without requiring the user to set aside time for dialogue.

[0009] This is a diagram illustrating the configuration of an information processing system according to one embodiment. This is a diagram illustrating the configuration of an information processing device according to one embodiment. This is a diagram illustrating the contents of the agent storage unit. This is a diagram illustrating the contents of the user storage unit. This is a diagram illustrating the configuration of a client terminal according to one embodiment. This is a flowchart illustrating the processing in the information processing device according to one embodiment. This is a diagram illustrating an example of the hardware configuration of the information processing device according to one embodiment. This is a diagram illustrating an example of the hardware configuration of a client terminal according to one embodiment.

[0010] Embodiments of the present invention will be described with reference to the attached drawings. The following embodiments are provided to facilitate understanding of the present invention and are not intended to limit its interpretation. Furthermore, the present invention can be modified in various ways without departing from its essence. Those skilled in the art can also adopt embodiments in which each of the elements described below is replaced with equivalent components, and such embodiments are also included within the scope of the present invention.

[0011] (System Configuration) An overview of this disclosure will be described using Figure 1. Figure 1 is a configuration diagram of an information processing system according to one embodiment. The information processing system comprises an information processing device 100 and a plurality of client terminals 200a, 200b, ..., 200n (hereinafter referred to as client terminals 200), and each client terminal 200 is connected to the information processing device 100 so as to be able to communicate with each other via a communication network N. The communication network N may be either a wired communication network or a wireless communication network composed of wired or wireless lines, and may be the Internet or a Local Area Network (LAN).

[0012] The information processing device 100 can have a first AI agent created based on first agent information and a second AI agent created based on second agent information interact for a predetermined purpose, obtain the evaluation results of the second AI agent on the first AI agent and / or the evaluation results of the first AI agent on the second AI agent based on the content of the interaction, and output matching results based on the evaluation results. The purpose of the interaction may include, but is not limited to, personnel recruitment and romantic matching.

[0013] The client terminal 200 is a terminal used by a user, and can be a smartphone, tablet, personal computer, etc. Here, users include job applicants, personnel recruiters, users of dating matching services, and providers of dating matching services. Users can provide agent information to the information processing device 100 via the client terminal 200, which is used to realize an AI agent. Users can also receive matching results from the information processing device 100 via the client terminal 200.

[0014] An AI agent may be a software program that uses generative AI. An AI agent is interactive software that, upon receiving instructions or requests in natural language, identifies the necessary processes to respond to those instructions or requests and automatically executes them. The generative AI may be, for example, a text generation AI, an image generation AI, or a multimodal generation AI.

[0015] (Functional Configuration: Information Processing Device) Figure 2 shows the configuration of an information processing device according to one embodiment. In Figure 2, only the necessary functional configuration is shown, assuming a single information processing device 100. However, the information processing device 100 can also be configured as part of a multifunctional distributed system consisting of multiple computer systems.

[0016] The information processing device 100 includes a storage unit 110, a site provision unit 120, a registration unit 130, an agent creation unit 140, a collection unit 150, a dialogue control unit 160, a result creation unit 170, and an output unit 180. Each of the units shown in Figure 2 can be realized, for example, by using a storage area or by having a processor execute a program stored in the storage area.

[0017] In this embodiment, it is desirable that the storage unit 110 includes an agent storage unit 111, a user storage unit 112, an extraction model 113, and a recommended method storage unit 114.

[0018] The agent memory unit 111 stores information about the agent. In one embodiment, it is desirable that the agent memory unit 111 includes the agent ID, agent name, dialogue purpose, evaluation points, and attributes, as shown in Figure 3. The agent ID is information that can uniquely identify the agent. The agent name is information that indicates the name of the agent used when outputting. The dialogue purpose stores information that indicates the purpose of the dialogue, such as "1: Job application," "2: Talent recruitment," or "3: Matching."

[0019] Evaluation points are information that indicates what is particularly emphasized when evaluating recruitment or matching. You can specify content that works in a positive or negative direction in the evaluation. Various information about the agent according to the purpose of the conversation is stored in the attributes. If the purpose of the conversation is "1: Job application," for example, information such as age, gender, education, work history, and selling points will be stored as attributes. If the purpose of the conversation is "2: Talent acquisition," for example, information such as company name, job title, and recruitment conditions will be stored as attributes. If the purpose of the conversation is "3: Matching," for example, information such as age, gender, education, work history, hobbies, and place of origin will be stored as attributes.

[0020] The "selling points" section contains information that applicants particularly want to highlight to recruiters. The "job title" section indicates the position being filled. As mentioned above, the agent information includes information that is necessary for interacting with other AI agents, such as the characteristics of the AI ​​agent that influence the content of the conversation and the mutual evaluation of the conversation, i.e., information equivalent to a person's personality, as well as information on attributes such as age and gender.

[0021] The user memory unit 112 stores user information. In one embodiment, it is desirable that the user memory unit 112 includes a user ID, purpose of use, agent ID, and email address, as shown in Figure 4. The user ID is information that can uniquely identify a user of the matching service site provided by the information processing device 100. The purpose of use stores information indicating the user's purpose for using the information processing system, such as "1: Job application," "2: Talent recruitment," "3: Matching participation - 1," "4: Matching participation - 2," and "5: Matching successful." Note that "3: Matching participation - 1" and "4: Matching participation - 2" can be used when dividing users participating in matching into two groups. For example, in the case of romantic matching, one of the pieces of information is stored according to gender. Also, in the case of matching entrepreneurs and investors, one of the pieces of information is stored according to their position.

[0022] The Agent ID stores information indicating the Agent ID of the Agent Memory Unit 111. For example, if the purpose of use is "1: Job Application," the Agent ID stores the Agent ID of the Agent Memory Unit 111 that identifies the agent corresponding to the user. If the purpose of use is "2: Talent Recruitment," the Agent ID stores the Agent ID of the Agent Memory Unit 111 that identifies the agent corresponding to the recruiter of the company to which the user belongs. If the purpose of use is "3: Matching Participation-1" or "4: Matching Participation-2," the Agent ID stores the Agent ID of the Agent Memory Unit 111 that identifies the agent corresponding to the user. Note that if the purpose of use is "5: Matching Successful," no information does not need to be stored in the Agent ID.

[0023] Extraction model 113 is a model that extracts information about behavior change techniques (BCTs) contained in the input data.

[0024] Here, the behavior change technique (BCT) according to this embodiment refers to a method or technique that acts upon (influences) behavior change. For example, according to BCTTv1 (Michie S, Richardson M, Johnston M, et al.: The behavior change technique taxonomy (v1) of 93 hierarchically clustered techniques: building an international consensus for the reporting of behavior change interventions. Ann Behav Med 2013; 46: 81-95.), 93 BCTs in 16 groups are defined. However, the definition of BCT is not limited to BCTTv1, and may be defined in any way as long as it comprehensively covers methods for behavior change.

[0025] The extraction model 113 may be generated by training it using data to which BCT annotation information has been added.

[0026] The recommended method memory unit 114 stores information on BCTs that are recommended to be included in the dialogue content. In one embodiment, it is desirable that the recommended method memory unit 114 includes the dialogue purpose and recommended BCTs. The dialogue purpose is the same information as the dialogue purpose in the agent memory unit 111. The recommended BCTs are information indicating BCTs that are recommended to be included in the dialogue content of the target dialogue purpose.

[0027] The site provision unit 120 provides the matching service site provided by the information processing device 100 on a network N, such as the internet. On this website, users can register and view data related to matching. To this end, the site provision unit 120 sends a web page written in HTML (HyperText Markup Language), etc., to the client terminal 200, and also receives user input results, etc., from the client terminal 200.

[0028] The site provider unit 120 creates a corresponding web page in response to user input and sends it to the client terminal 200. Specifically, the site provider unit 120 sends various web pages to the client terminal 200, such as a web page for the user to register their information and / or agent information, a web page for the user to select an agent with whom they can interact, and a web page for the user to confirm the matching results.

[0029] The registration unit 130 registers user information and / or agent information. In this embodiment, when the registration unit 130 receives a registration request from the client terminal 200, it generates a user ID and / or agent ID and registers them in the user storage unit 112 and / or agent storage unit 111 based on the information included in the registration request. In one embodiment, the client terminal 200 sends a registration request that includes agent information in addition to the user's purpose of use. For example, it is assumed that a user applying for a job sends a registration request via the client terminal 200 that includes the purpose of use "1: Job application" in addition to the agent name and attributes as information about the agent corresponding to the user.

[0030] In one embodiment, the client terminal 200 transmits separately a registration request containing user information and a registration request for registering agent information. For example, it is assumed that a user who is a matching service provider first transmits a registration request via the client terminal 200 that includes the purpose of use "5: Matching successful," and then transmits a registration request that includes the agent name, dialogue purpose "3: Matching," and attributes as information for the agent corresponding to the matching participant.

[0031] The agent creation unit 140 creates an AI agent that mimics a desired user based on the agent information. In this embodiment, when a dialogue request is received from the client terminal 200, at least two or more AI agents are created based on the information contained in the dialogue request. In one example, the dialogue request includes a user ID, an agent ID, etc. For example, if user X, who is applying for a job, wishes to have an interview with a recruiter at company A from the list of available agents, it is assumed that a dialogue request is sent from the client terminal 200 that includes a user ID corresponding to user X and an agent ID corresponding to the recruiter at company A.

[0032] In another example, the dialogue request may include a user ID, agent conditions, etc. For instance, if female user F, who is participating in the matching process, wishes to be matched with a male agent, it is assumed that a dialogue request including the user ID corresponding to user F and the agent condition "30 years of age or older" will be sent from the client terminal 200.

[0033] In another example, a dialogue request may include a user ID, dialogue purpose, and agent keywords. For instance, if user X, who is applying for a job, wishes to interview with a recruiter at company E, which is not on the list of available agents, it is expected that a dialogue request will be sent from client terminal 200, including the user ID corresponding to user X, dialogue purpose "2: Recruitment," and agent keywords "Company E, Sales Position."

[0034] The AI ​​agent may be configured within the information processing device 100, or it may be configured within an external information processing device.

[0035] The collection unit 150 collects agent information. In this embodiment, when the collection unit 150 receives a dialogue request containing an agent keyword from the client terminal 200, it performs an internet search based on the dialogue purpose and agent keyword included in the dialogue request, and registers the agent information in the agent storage unit 111 based on the information collected from the internet.

[0036] The dialogue control unit 160 causes the first AI agent and the second AI agent to converse for a predetermined purpose. For example, when causing an applicant agent with dialogue purpose "1: Job application" to converse with a recruiting agent with dialogue content "2: Talent acquisition", the dialogue control unit 160 1) obtains question information to be asked of the applicant agent from the recruiting agent, 2) inputs the question information into the applicant agent and obtains the answer information, and 3) inputs the question information and answer information into the recruiting agent and obtains the evaluation result for the applicant agent.

[0037] The evaluation results may be expressed as numerical values, or as rankings or other evaluation results expressed by letters or symbols such as A, B, C, or Class A, B, or C. In one embodiment, the dialogue control unit 160 acquires evaluation results that reflect the conscious load.

[0038] Here, the conscious load is a value corresponding to the mental burden and attentional resources imposed on a user processing predetermined information. In this embodiment, it is a value corresponding to the mental burden and attentional resources that the user modeled by the AI ​​agent is expected to experience when processing information related to the dialogue. The conscious load can be imposed on both the agent asking the question and the agent receiving the question and providing the answer. For example, the agent asking the question incurs a conscious load to generate the question information, and the agent receiving the question and providing the answer incurs a conscious load to understand the question information and generate the answer information. Furthermore, the agent asking the question incurs a conscious load to understand and evaluate the answer information.

[0039] Furthermore, when arranging a dialogue between agent 1 and agent 2, both with dialogue objective "3: matching", the dialogue control unit 160 performs the following actions: 1) obtains question information 1 from agent 2 to ask agent 1; 2) inputs question information 1 to agent 1 and obtains answer information 1; 3) inputs question information 1 and answer information 1 to agent 2 and obtains evaluation result 1 for agent 1; 4) obtains question information 2 from agent 1 to ask agent 2; 5) inputs question information 2 to agent 2 and obtains answer information 2; 6) inputs question information 2 and answer information 2 to agent 1 and obtains evaluation result 2 for agent 2.

[0040] The dialogue control unit 160 can generate input information for the agent using the evaluation points and attributes of the agent memory unit 111. For example, when obtaining question information from the recruiting agent, the dialogue control unit 160 can generate input information using the evaluation points and obtain question information that takes the evaluation points into consideration. Also, when obtaining evaluation results for the applicant agent, the dialogue control unit 160 can generate input information using the evaluation points and obtain evaluation results that take the evaluation points into consideration.

[0041] The result creation unit 170 creates matching results based on the evaluation results obtained by the dialogue control unit 160. For example, if the dialogue control unit 160 obtains evaluation results for an applicant agent, the result creation unit 170 creates matching results for users related to the recruiting agent and / or matching results for users related to the applicant agent.

[0042] In one embodiment, the result creation unit 170 may input question information and answer information into the extraction model 113 to extract BCT information contained in the input data. If BCT information contained in the input data is extracted, the result creation unit 170 may create a matching result that includes the extracted BCT information.

[0043] Also, in one embodiment, the result creation unit 170 may refer to the recommended method storage unit 114 based on the dialogue purpose and acquire information on BCTs that are recommended to be included in the dialogue content. When information on BCTs that are recommended to be included in the dialogue content is acquired, the result creation unit 170 may create a matching result including the acquired BCT information. In addition, when information on BCTs included in the input data is also extracted, the result creation unit 170 may generate a matching result including difference information between the BCT information included in the input data and the BCT information that is recommended to be included in the dialogue content.

[0044] The output unit 180 outputs the matching result. In this embodiment, the output unit 180 transmits the matching result created by the result creation unit 170 to the desired client terminal 200.

[0045] (Functional Configuration: Client Terminal)

[0046] FIG. 5 is a diagram showing the configuration of a client terminal according to an embodiment. The client terminal 200 includes a storage unit 210, a registration request unit 220, a dialogue request unit 230, and a display unit 240. Each unit shown in FIG. 5 can be realized, for example, by using a storage area or by the processor executing a program stored in the storage area. In this embodiment, the registration request unit 220, the dialogue request unit 230, and the display unit 240 can be realized by a browser realized by the processor executing a program stored in the storage area. However, in other embodiments, they may be realized by executing an application for a matching service stored in the storage area of the client terminal 200.

[0047] The registration request unit 220 transmits a registration request for a user and / or an agent. In one embodiment, the registration request unit 220 transmits a registration request including the purpose of use, agent information, etc. to the information processing device 100. In another embodiment, the registration request unit 220 separately transmits a registration request including the purpose of use and a registration request including agent information to the information processing device 100.

[0048] The dialogue request unit 230 transmits a dialogue request specifying multiple agents that wish to engage in dialogue. In one embodiment, the dialogue request unit 230 transmits a dialogue request including a user ID and an incoming agent ID to the information processing device 100. In another embodiment, the dialogue request unit 230 transmits a dialogue request including a user ID and agent conditions to the information processing device 100. In yet another embodiment, the dialogue request unit 230 transmits a dialogue request including a user ID, dialogue purpose, and agent keywords to the information processing device 100.

[0049] The display unit 240 receives the matching result as a response to the dialogue request and displays it on the display device of the client terminal 200.

[0050] (Operation) Referring to Figure 6, the processing in the information processing device 100 according to one embodiment will be described in detail. In this embodiment, it is assumed that various data are stored in the agent storage unit 111, extraction model 113, and recommended method storage unit 114 of the storage unit 110 before the processing in Figure 6 is started.

[0051] (Example 1) In this example, we will describe a case in which user X, who is applying for a job, wishes to have an interview with the hiring manager of company A from the list of available agents.

[0052] In step S101, the site provision unit 120 of the information processing device 100 sends a web page to the client terminal 200 for registering user information and / or agent information.

[0053] In response to user X's actions in the browser, when a registration request is sent from the client terminal 200 (step S102), in step S103, the registration unit 130 of the information processing device 100 generates a user ID and / or agent ID and registers them in the user storage unit 112 and / or agent storage unit 111 based on the information included in the registration request. Here, it is assumed that the client terminal 200 receives a registration request including the purpose of use "1: Job application", agent name "Applicant X", and attributes, and the registration unit 130 generates a user ID "XXXXX" and agent ID "U001", and registers the record in the user storage unit 112 and the record in the agent storage unit 111.

[0054] In step S104, the site provision unit 120 sends a web page to the client terminal 200 for the user to select an agent they can interview.

[0055] In response to user X's operation in the browser, when a dialogue request is sent from the client terminal 200 (step S105), in step S106, the agent creation unit 140 of the information processing device 100 creates at least two AI agents based on the information contained in the dialogue request. Here, it is assumed that the agent creation unit 140 receives a dialogue request from the client terminal 200 including user ID "XXXXX" and agent ID "C001", and, referring to the agent storage unit 111 shown in Figure 3(A), creates an applicant agent based on the agent information of agent ID "U001" associated with user ID "XXXXX", and a recruitment agent based on the agent information of agent ID "C001".

[0056] In step S107, the dialogue control unit 160 of the information processing device 100 causes the first AI agent and the second AI agent to converse for a predetermined purpose. Here, the dialogue control unit 160 sequentially performs the following processes: 1) obtains question information to be asked of the applicant agent from the hiring agent, 2) inputs the question information into the applicant agent and obtains answer information, and 3) inputs the question information and answer information into the hiring agent and obtains an evaluation result for the applicant agent.

[0057] As described above, in one embodiment, the dialogue control unit 160 acquires evaluation results that reflect the conscious load. Here, the dialogue control unit 160 acquires evaluation results that reflect the conscious load on the recruiting agent in order to generate question information to ask the applicant agent, and the conscious load on the recruiting agent in order to evaluate the applicant agent based on the answer information.

[0058] For example, the fewer resources the recruiting agent consults to generate the question information, the lower the cognitive load in the evaluation result. Similarly, the fewer resources the recruiting agent consults to understand the answer information, the lower the cognitive load in the evaluation result. Furthermore, the fewer the number of interactions until the answer information is obtained, the lower the cognitive load in the evaluation result. For example, if the applicant agent cannot understand the question information and the answer to the first question is not the intended answer, the dialogue control unit 160 may input the first question information and the first answer information to the recruiting agent to obtain a second question information by rephrasing the first question information, and input the second question information to the applicant agent to obtain a second answer information. In this way, the dialogue control unit 160 obtains rephrased question information until the answer information to the question information becomes the intended answer information, and the fewer the number of interactions until the intended answer information is obtained, the lower the cognitive load in the evaluation result.

[0059] In step S108, the result creation unit 170 of the information processing device 100 creates matching results based on the evaluation results acquired by the dialogue control unit 160. Here, the result creation unit 170 creates matching results for users related to the recruiting agent and matching results for users related to the applicant agent, based on the evaluation results for the applicant agent acquired by the dialogue control unit 160.

[0060] In one embodiment, the result creation unit 170 may input question information and answer information into the extraction model 113 to extract BCT information contained in the input data. If BCT information contained in the input data is extracted, the result creation unit 170 may create a matching result that includes the extracted BCT information.

[0061] In one embodiment, the result creation unit 170 may also refer to the recommended method storage unit 114 based on the dialogue objective to obtain information on BCTs that are recommended to be included in the dialogue content. If information on BCTs that are recommended to be included in the dialogue content is obtained, the result creation unit 170 may create a matching result that includes the obtained BCT information. If information on BCTs included in the input data is also extracted, the result creation unit 170 may generate a matching result that includes difference information between the BCT information included in the input data and the BCT information that is recommended to be included in the dialogue content.

[0062] In step S109, the site provision unit 120 sends a web page to the client terminal 200 for confirmation of the matching results. Here, it is assumed that the site provision unit 120 sends a web page of matching results for users related to the applicant agent to the client terminal 200 of user X.

[0063] In step S110, the output unit 180 of the information processing device 100 transmits the matching result to the client terminal 200. Here, the output unit 180 sends the matching result, which is intended for users associated with the recruitment agent, to the email address of the recruitment officer at company A. It is desirable that the transmission of the matching result to the recruitment officer at company A, who was selected as the dialogue partner, rather than to user X who sent the dialogue request, be done with the consent of user X who sent the dialogue request.

[0064] (Additional Embodiment: Dialogue Environment Settings) In Embodiment 1 described above, an example was given in which an applicant agent and a recruiting agent interacted with only the dialogue purpose specified. In another embodiment, in addition to the dialogue purpose, other dialogue environment settings can be specified in which the applicant agent and the recruiting agent can interact.

[0065] The process proceeds in the same manner as in Embodiment 1 from steps S101 to S104 in Figure 6. In this embodiment, for example, in step S104 of Figure 6, on the web page for selecting the agent that was sent, user X can specify the dialogue environment settings. For example, the dialogue environment settings include the dialogue date, dialogue start time, dialogue language, etc.

[0066] When user X specifies the conversation environment settings in the browser, a conversation request including the conversation environment settings is sent from the client terminal 200 (step S105). Here, it is assumed that a conversation request including the user ID "XXXXX", agent ID "C001", interview start time "13:00", and conversation language "Japanese" corresponding to user X is sent from the client terminal 200.

[0067] Similar to the first embodiment described above, in step S106, the agent creation unit 140 of the information processing device 100 creates at least two AI agents based on the information included in the dialogue request. Here, the agent creation unit 140 refers to the agent storage unit 111 shown in Figure 3(A) and creates an applicant agent based on the information of the agent with agent ID "U001" associated with user ID "XXXXX", and a recruitment agent based on the information of the agent with agent ID "C001".

[0068] In step S107, the dialogue control unit 160 of the information processing device 100 causes the first AI agent and the second AI agent to converse for a predetermined purpose. Here, the dialogue control unit 160 sequentially performs the following processes: 1) obtains question information to be asked of the applicant agent from the hiring agent, 2) inputs the question information into the applicant agent and obtains answer information, and 3) inputs the question information and answer information into the hiring agent and obtains an evaluation result for the applicant agent.

[0069] As described above, the dialogue control unit 160 acquires evaluation results that reflect the conscious load. In this embodiment, in addition to the number of resources referenced and the number of dialogues required to obtain the answer information, as described in Embodiment 1, or instead of the number of resources referenced and the number of dialogues required to obtain the answer information, the conscious load according to the dialogue environment settings may be reflected in the evaluation results.

[0070] Specifically, the dialogue control unit 160 can generate input information for the agent using the dialogue environment settings in addition to the evaluation points and attributes of the agent memory unit 111. For example, when obtaining question information from the recruiting agent, the dialogue control unit 160 can generate input information using the dialogue environment settings and obtain question information that takes the dialogue environment settings into consideration. Also, when obtaining evaluation results for applicant agents, the dialogue control unit 160 can generate input information using the dialogue environment settings and obtain evaluation results that take the dialogue environment settings into consideration.

[0071] For example, a recruiting agent might give different evaluation results depending on the interview start time, even when interacting with the same applicant agent. For instance, a recruiting agent who tends to communicate with others in the morning and focus on their own work in the afternoon might receive a lower evaluation result for interviews starting in the morning than in the afternoon.

[0072] Furthermore, in other examples, a recruiting agent may return different evaluation results depending on the language of communication, even when interacting with the same applicant agent. For instance, a recruiting agent who is more proficient in English than Japanese may obtain evaluation results indicating a lower mental burden when communicating in English than when communicating in Japanese.

[0073] The process from step S108 onward is the same as in the first embodiment described above, so a detailed explanation will be omitted.

[0074] (Additional Embodiments) In the above Embodiment 1, we described an example in which user X, who applies for a job, selects a recruiter from company A from a list of available agents for interviews. In another embodiment, user X, who applies for a job, can request an interview with a recruiter from company E, which is not on the list of available agents for interviews.

[0075] Steps S101 to S104 in Figure 6 proceed in the same manner as in the first embodiment described above. In step S105, for example, in response to user X's operation on the web page for selecting an agent, which was sent in step S104 of Figure 6, the client terminal 200 sends a dialogue request requesting an interview with an agent created based on the dialogue purpose "2: Recruitment" and agent keyword "Company E, Sales Position," rather than an agent in the list of available agents. Here, it is assumed that the client terminal 200 sends a dialogue request including user ID "XXXXX" corresponding to user X, dialogue purpose "2: Recruitment," and agent keyword "Company E, Sales Position."

[0076] In step S106, prior to processing by the agent creation unit 140, the collection unit 150 of the information processing device 100 collects agent information. In this embodiment, the collection unit 150 performs an internet search based on the dialogue purpose and agent keyword included in the dialogue request, and registers the agent information in the agent storage unit 111 based on the information collected from the internet. Here, it is assumed that the collection unit 150 performs an internet search based on the dialogue purpose "2: Recruitment" and the agent keyword "Company E, Sales Position," and registers the agent information in the agent storage unit 111 based on the information collected from the internet.

[0077] Next, the agent creation unit 140 creates AI agents. Here, the agent creation unit 140 refers to the agent storage unit 111 shown in Figure 3(A) and creates an applicant agent based on the agent information of agent ID "U001" associated with user ID "XXXXX", and a recruitment agent based on the agent information of agent ID "A001".

[0078] The processing from step S107 onward is the same as in the first embodiment described above, so a detailed explanation will be omitted. In this embodiment, the agent selected as the conversation partner is not created based on information in the agent storage unit 111 that has been registered in advance in relation to the user who wishes to have a conversation, but rather based on information collected based on predetermined conditions. Therefore, the processing in step S110 is skipped, and the information processing device 100 terminates processing.

[0079] (Example 2) This example describes a case in which a female user F participating in a dating matching service wishes to be matched with a male agent.

[0080] In step S101, the site provision unit 120 of the information processing device 100 sends a web page to the client terminal 200 for registering user information and / or agent information.

[0081] In response to user F's operation in the browser, when a registration request is sent from the client terminal 200 (step S102), in step S103, the registration unit 130 of the information processing device 100 generates a user ID and / or agent ID and registers them in the user storage unit 112 and / or agent storage unit 111 based on the information included in the registration request. Here, it is assumed that the registration unit 130 receives a registration request from the client terminal 200 including the purpose of use "3: Matching Participation-1", the agent name "Female Agent 1", and attributes, and generates a user ID "F0101" and an agent ID "U101", and registers the record in the user storage unit 112 and the record in the agent storage unit 111.

[0082] In step S104, the site provision unit 120 sends a web page to the client terminal 200 for selecting an agent to perform matching.

[0083] In response to user F's operation in the browser, when a dialogue request is sent from the client terminal 200 (step S105), in step S106, the agent creation unit 140 of the information processing device 100 creates at least two AI agents based on the information contained in the dialogue request. Here, it is assumed that the client terminal 200 receives a dialogue request including user ID "F0101" and agent condition "30 years of age or older," and the agent creation unit 140, referring to the agent storage unit 111 shown in Figure 3(B), creates female agent 1 based on the agent information of agent ID "U101" associated with user ID "F0101," and male agents 1 to 10 whose purpose of use "4: Matching Participation-2" satisfies the agent condition "30 years of age or older."

[0084] In step S107, the dialogue control unit 160 of the information processing device 100 has the first AI agent and the second AI agent converse for a predetermined purpose. In this case, since a total of 10 male agents have been created, the dialogue control unit 160 will acquire each male agent one by one and perform the following processing. Specifically, the dialogue control unit 160 will sequentially perform the following processing: 1) acquire question information 1 to ask female agent 1 from the male agent, 2) input question information 1 to female agent 1 and acquire answer information 1, 3) input question information 1 and answer information 1 to the male agent and acquire evaluation result 1 for female agent 1, 4) acquire question information 2 to ask the male agent from female agent 1, 5) input question information 2 to the male agent and acquire answer information 2, and 6) input question information 2 and answer information 2 to female agent 1 and acquire evaluation result 2 for the male agent.

[0085] As described above, in one embodiment, the dialogue control unit 160 acquires an evaluation result that reflects the conscious load. Here, the dialogue control unit 160 acquires an evaluation result 1 for the female agent 1 that reflects the conscious load on the male agent to generate question information 1 to ask the female agent 1, and the conscious load on the male agent to evaluate the female agent 1 based on the answer information 1. Similarly, the dialogue control unit 160 acquires an evaluation result 2 for the male agent that reflects the conscious load on the female agent 1 to generate question information 2 to ask the male agent, and the conscious load on the female agent 1 to evaluate the male agent based on the answer information 2.

[0086] In one example, as described in Example 1, the evaluation result may reflect the mental load based on the number of resources referenced and the number of dialogues required to obtain the answer information. In another example, the closer the number of characters in the question information and the answer information, the lower the evaluation result in terms of mental load. In this example, it is assumed that the closer the number of characters, the better the pace of the dialogue and the lower the mental load. Conversely, if the number of characters in one is significantly larger than the other, it is assumed that the pace of the dialogue is poor and the mental load is high.

[0087] Furthermore, in one example, the more similar the attributes of female agent 1 and male agent 1 are, the lower the cognitive load will be in the evaluation results. For example, the closer their age and educational level, the lower the cognitive load will be in the evaluation results. Also, if they have the same hobbies or hometown, the cognitive load will be lower compared to when their hobbies or hometowns are different.

[0088] In step S108, the result creation unit 170 of the information processing device 100 creates matching results based on the evaluation results acquired by the dialogue control unit 160. Here, the result creation unit 170 creates matching results for user F related to female agent 1 based on the evaluation results for female agent 1 and / or evaluation results 2 for male agent acquired by the dialogue control unit 160.

[0089] The result creation unit 170 may include in the matching results an evaluation result in which the evaluation of female agent 1 is higher than a predetermined standard, or an evaluation result in which the evaluation of male agent is higher than a predetermined standard, or an evaluation result in which the evaluation of female agent 1 is higher than a predetermined standard and the evaluation of male agent is higher than a predetermined standard.

[0090] In step S109, the site provision unit 120 sends a web page to the client terminal 200 for confirmation of the matching results. Here, it is assumed that the site provision unit 120 sends a web page of matching results for users related to female agent 1 to the client terminal 200 of user F. In this embodiment, the processing in step S110 is skipped and the information processing device 100 terminates processing.

[0091] (Additional Embodiment: Dialogue Environment Settings) In the above embodiment 2, an example was described in which female agent 1 and male agent are made to interact with only the dialogue purpose specified. In another embodiment, female agent 1 and male agent can be made to interact with other dialogue environment settings specified in addition to the dialogue purpose.

[0092] The process proceeds in the same manner as in the above-described embodiment 2, from steps S101 to S104 in Figure 6. In this embodiment, for example, in step S104 of Figure 6, on the web page for selecting the agent to perform matching, user F can specify the dialogue environment settings. For example, the dialogue environment settings include the dialogue date, dialogue start time, dialogue method, dialogue schedule, etc. The dialogue method specifies information indicating the means of dialogue, such as "1: direct dialogue", "2: telephone", "3: email", etc. The dialogue schedule specifies information indicating the dialogue schedule from the start of the dialogue until the evaluation result is obtained, such as "1: once", "2: daily / week", "3: once a week / month", etc.

[0093] When user F specifies the conversation environment settings in the browser, a conversation request including the conversation environment settings is sent from the client terminal 200 (step S105). Here, it is assumed that a conversation request including user ID "F0101", agent condition "30 years of age or older", conversation start time "19:00", conversation method "3: email", and conversation schedule "2: daily / weekly" is sent from the client terminal 200.

[0094] Similar to the above embodiment 2, in step S106, the agent creation unit 140 of the information processing device 100 creates at least two AI agents based on the information included in the dialogue request. Here, the agent creation unit 140 refers to the agent storage unit 111 shown in Figure 3(B) and creates a female agent 1 based on the agent information of agent ID "U101" associated with user ID "F0101", and male agents 1 to 10 whose purpose of use "4: Matching Participation-2" satisfies the agent condition "30 years of age or older".

[0095] In step S107, the dialogue control unit 160 of the information processing device 100 has the first AI agent and the second AI agent converse for a predetermined purpose. In this case, since a total of 10 male agents have been created, the dialogue control unit 160 will acquire each male agent one by one and perform the following processing. Specifically, the dialogue control unit 160 will sequentially perform the following processing: 1) acquire question information 1 to ask female agent 1 from the male agent, 2) input question information 1 to female agent 1 and acquire answer information 1, 3) input question information 1 and answer information 1 to the male agent and acquire evaluation result 1 for female agent 1, 4) acquire question information 2 to ask the male agent from female agent 1, 5) input question information 2 to the male agent and acquire answer information 2, and 6) input question information 2 and answer information 2 to female agent 1 and acquire evaluation result 2 for the male agent.

[0096] As described above, the dialogue control unit 160 acquires evaluation results that reflect the conscious load. In this embodiment, in addition to the number of resources referenced and the number of dialogues required to obtain answer information, as described in Example 1, and the tempo of the dialogue and the similarity of agent attributes based on the number of characters in the question information and the number of characters in the answer information, as described in Example 2, or in lieu thereof, the conscious load according to the dialogue environment settings may be reflected in the evaluation results. For example, even when a male agent is conversing with the same female agent 1, it is conceivable that different evaluation results will be returned depending on the dialogue start time, dialogue means, and dialogue schedule. Similarly, even when a female agent 1 is conversing with the same male agent, it is conceivable that different evaluation results will be returned depending on the dialogue start time, dialogue means, and dialogue schedule.

[0097] For example, male agents who tend to prefer direct conversation as a means of communication will evaluate the mental burden as lower when the means of communication is direct conversation rather than by phone or email. Also, for example, male agents who tend to prefer frequent conversations with the opposite sex will evaluate the mental burden as lower when the conversation schedule is daily rather than once a week.

[0098] The process from step S108 onward is the same as in the above embodiment 2, so a detailed explanation will be omitted.

[0099] (Example 3) In this example, we will describe a case in which User Z, a provider of a dating matching service, desires to obtain pairs with a high matching rate.

[0100] In step S101, the site provision unit 120 of the information processing device 100 sends a web page to the client terminal 200 for registering user information and / or agent information.

[0101] In response to user Z's actions in the browser, when a registration request is sent from the client terminal 200 (step S102), in step S103, the registration unit 130 of the information processing device 100 generates a user ID and / or agent ID and registers them in the user storage unit 112 and / or agent storage unit 111 based on the information included in the registration request. Here, it is assumed that the client terminal 200 has received a registration request including the purpose of use "5: Matching successful," and the registration unit 130 has generated a user ID "ZZZZZ" and registered the record in the user storage unit 112.

[0102] In step S104, the site provision unit 120 sends a web page to the client terminal 200 for selecting an agent to perform matching.

[0103] In response to user Z's actions in the browser, when a dialogue request is sent from the client terminal 200 (step S105), in step S106, the agent creation unit 140 of the information processing device 100 creates at least two AI agents based on the information contained in the dialogue request. Here, it is assumed that the client terminal 200 receives a dialogue request including user ID "ZZZZZ" and agent condition "not specified," and the agent creation unit 140, referring to the agent storage unit 111 shown in Figure 3(B), creates a plurality of female agents 1 to N and a plurality of male agents 1 to M.

[0104] In step S107, the dialogue control unit 160 of the information processing device 100 has the first AI agent and the second AI agent converse for a predetermined purpose. In this case, a total of N female agents and a total of M male agents have been created, so the dialogue control unit 160 will acquire the female agent and the male agent one by one and perform the following processing. Specifically, the dialogue control unit 160 will sequentially perform the following processing: 1) acquire question information 1 to ask the female agent from the male agent, 2) input question information 1 to the female agent and acquire answer information 1, 3) input question information 1 and answer information 1 to the male agent and acquire evaluation result 1 for the female agent, 4) acquire question information 2 to ask the male agent from the female agent, 5) input question information 2 to the male agent and acquire answer information 2, and 6) input question information 2 and answer information 2 to the female agent and acquire evaluation result 2 for the male agent. As described above, in one embodiment, the dialogue control unit 160 acquires evaluation results that reflect the conscious load.

[0105] In step S108, the result creation unit 170 of the information processing device 100 creates a matching result based on the evaluation results acquired by the dialogue control unit 160. Here, the result creation unit 170 creates a matching result for user Z based on the evaluation results for the female agent and the evaluation results 2 for the male agent acquired by the dialogue control unit 160.

[0106] The result creation unit 170 may include in the matching results all pairs of female agents and male agents whose evaluation of each other is higher than a predetermined standard, or it may include in the matching results the top 10 pairs of pairs whose evaluations of each other are higher than a predetermined standard.

[0107] In step S109, the site provision unit 120 sends a web page to the client terminal 200 for confirmation of the matching results. Here, it is assumed that the site provision unit 120 sends the matching result web page to user Z's client terminal 200. In this embodiment, the processing in step S110 is skipped, and the information processing device 100 terminates processing.

[0108] (Hardware Configuration: Information Processing Device) Next, an example of the hardware configuration of the information processing device 100 will be described with reference to Figure 7. As shown in Figure 7, the information processing device 100 includes, for example, a processor 101, memory 102, storage device 103, input I / F unit 104, data I / F unit 105, communication I / F unit 106, and display device 107.

[0109] The processor 101 is a control unit that controls various processes in the information processing device 100 by executing a program stored in the memory 102.

[0110] Memory 102 is a storage medium such as RAM (Random Access Memory). Memory 102 temporarily stores the program code of the program executed by the processor 101, as well as data required during program execution.

[0111] The storage device 103 is a non-volatile storage medium such as a hard disk drive (HDD) or flash memory. The storage device 103 stores the operating system and various programs for realizing the above configurations.

[0112] The input interface unit 104 is a device for receiving input from the user. The input interface unit 104 may be, for example, a keyboard, mouse, touch panel, various sensors, or a wearable device. The input interface unit 104 may be connected to the information processing device 100 via an interface such as USB (Universal Serial Bus).

[0113] The data I / F unit 105 is a device for inputting data from outside the information processing device 100. The data I / F unit 105 is, for example, a drive device for reading data stored on various storage media. The data I / F unit 105 may be provided outside the information processing device 100. If the data I / F unit 105 is provided outside the information processing device 100, it is connected to the information processing device 100 via an interface such as USB.

[0114] The communication interface unit 106 is a device for performing data communication with an external device of the information processing device 100 via a network such as the Internet, either by wire or wireless connection. The communication interface unit 106 may be located outside the information processing device 100. If the communication interface unit 106 is located outside the information processing device 100, it is connected to the information processing device 100 via an interface such as USB.

[0115] The display device 107 is a device for displaying various types of information. The display device 107 may be, for example, a liquid crystal display, an organic EL (Electro-Luminescence) display, or a display for a wearable device. The display device 107 may be provided outside the information processing device 100. If the display device 107 is provided outside the information processing device 100, it is connected to the information processing device 100, for example, via a display cable. Also, if a touch panel is used as the input I / F unit 104, the display device 107 may be configured to be integrated with the input I / F unit 104.

[0116] (Hardware Configuration: Client Terminal) Next, an example of the hardware configuration of the client terminal 200 will be described with reference to Figure 8. As shown in Figure 8, the client terminal 200 includes, for example, a processor 201, memory 202, storage device 203, input I / F unit 204, data I / F unit 205, communication I / F unit 206, and display device 207.

[0117] Note that the processor 201, memory 202, storage device 203, input I / F unit 204, data I / F unit 205, communication I / F unit 206, and display device 207 have the same configuration as the processor 101, memory 102, storage device 103, input I / F unit 104, data I / F unit 105, communication I / F unit 106, and display device 107, respectively, so a detailed explanation is omitted.

[0118] As described above, according to one embodiment, the information processing device 100 has a first AI agent created based on first agent information and a second AI agent created based on second agent information interact for a predetermined purpose, and acquires and outputs the first evaluation result of the second AI agent with respect to the first AI agent and / or the second evaluation result of the first AI agent with respect to the second AI agent. This allows the information processing device 100 to provide the evaluation results of other users with respect to the user and / or the evaluation results of the user with respect to other users without requiring the user to allocate time for interaction. Specifically, the interaction between the first AI agent, which mimics the personality, behavior, and habits of the user, and the AI ​​agent, which mimics the personality, behavior, and habits of other users, will mimic the interaction between the user and other users. As a result, the evaluation results between AI agents can be provided as evaluation results between users.

[0119] Furthermore, according to one embodiment, the information processing device 100 has a first AI agent created based on first agent information and a second AI agent created based on second agent information interact for a predetermined purpose, and obtains and outputs a first evaluation result for the first AI agent that reflects the consciousness load of the second AI agent and / or a second evaluation result for the second AI agent that reflects the consciousness load of the first AI agent. This allows the information processing device 100 to provide the user with evaluation results from other users and / or evaluation results of the user from other users without requiring the user to allocate time for interaction. Specifically, the interaction between the first AI agent, which mimics the personality, behavior, and habits of the user, and the AI ​​agent, which mimics the personality, behavior, and habits of other users, will mimic the interaction between the user and other users. As a result, the evaluation results between AI agents can be provided as evaluation results between users.

[0120] The information processing device 100 receives a user ID that specifies the first agent information and the purpose of the dialogue, and an agent ID that specifies the second agent information, and can then cause the first AI agent and the second AI agent to interact based on the specified information.

[0121] Furthermore, the information processing device 100 can receive a keyword that specifies second agent information and create a second AI agent using the information collected based on that keyword, and the user can obtain evaluation results from dialogue with the AI ​​agent assumed by the predetermined keyword.

[0122] Furthermore, by reflecting the conscious load of the AI ​​agent in the acquisition of evaluation results, it is expected that the evaluation results between AI agents can be brought closer to the evaluation results between users. Specifically, the information processing device 100 can acquire evaluation results that reflect values ​​corresponding to the mental load and attentional resources that each AI agent is expected to incur when processing information related to the conversation.

[0123] Furthermore, by enabling the creation of multiple second AI agents, the information processing device 100 can provide evaluation results for each second AI agent against the first AI agent in a one-to-many combination. In addition, the information processing device 100 can receive conditions that specify second agent information and create multiple second AI agents based on those conditions, thereby enabling the first AI agent to interact with a second AI agent that meets the conditions.

[0124] Furthermore, the information processing device 100 can obtain an evaluation that reflects the attentional load of the AI ​​agent according to the attentional resources by acquiring an evaluation result that indicates a lower attentional load the less resources each AI agent has referenced.

[0125] Furthermore, the information processing device 100 can obtain an evaluation that reflects the conscious load of the AI ​​agent according to the mental burden by acquiring an evaluation result that indicates a lower conscious load the fewer interactions are required to obtain predetermined response information.

[0126] Furthermore, the information processing device 100 can obtain different evaluation results depending on the dialogue environment settings, even for dialogues between the same AI agents, by acquiring evaluation results that reflect the conscious load according to the dialogue environment settings, which include at least one of the dialogue date, dialogue start time, dialogue language, dialogue means, and dialogue schedule.

[0127] Furthermore, the information processing device 100 can obtain an evaluation that reflects the conscious load of the AI ​​agent in relation to their mental burden by acquiring an evaluation result that indicates a lower conscious load when the number of characters in the question information and the number of characters in the answer information are close together.

[0128] Furthermore, by enabling the creation of multiple first AI agents and multiple second AI agents, the information processing device 100 can provide evaluation results for each second AI agent against each first AI agent in an N-to-M combination.

[0129] Furthermore, the information processing device 100 can provide the user with better combinations by creating matching results based on evaluation results that are higher than predetermined standards.

[0130] Furthermore, the information processing device 100 can extract behavioral change methods contained in the dialogue content and create matching results that include the extracted behavioral change method information, thereby providing the user with matching results analyzed based on behavioral science theory.

[0131] Furthermore, the information processing device 100 can provide the user with effective feedback based on behavioral science theory, while reflecting the user's current situation, by creating a matching result that includes difference information between the behavioral change methods included in the dialogue content and the behavioral change methods that are recommended to be included in the dialogue content.

[0132] 100 Information processing device, 101 Processor, 102 Memory, 103 Storage device, 104 Input I / F unit, 105 Data I / F unit, 106 Communication I / F unit, 107 Display device, 110 Storage unit, 111 Agent storage unit, 112 User storage unit, 113 Extraction model, 114 Recommended method storage unit, 120 Site provision unit (output unit), 130 Registration unit, 140 Agent creation unit (receiving unit), 150 Collection unit, 160 Dialogue control unit, 170 Result creation unit, 180 Output unit, 200 Client terminal, 201 Processor, 202 Memory, 203 Storage device, 204 Input I / F unit, 205 Data I / F unit, 206 Communication I / F unit, 207 Display device, 210 Storage unit, 220 Registration request unit, 230 Dialogue request unit, 240 Display unit, N Network

Claims

1. An information processing device comprising: an agent creation unit that creates an AI agent based on agent information; a dialogue control unit that causes a first AI agent created based on first agent information and a second AI agent created based on second agent information to interact for a predetermined purpose, and obtains a first evaluation result of the second AI agent with respect to the first AI agent and / or a second evaluation result of the first AI agent with respect to the second AI agent; a result creation unit that creates a matching result based on the first evaluation result and / or the second evaluation result; and an output unit that outputs the matching result.

2. The information processing apparatus according to claim 1, wherein the result creation unit extracts a first behavioral modification method included in the content of the dialogue between the first AI agent and the second AI agent, and creates a matching result including information on the first behavioral modification method.

3. The information processing apparatus according to claim 1, further comprising: a receiving unit that receives information specifying the first agent information, information specifying the second agent information including keywords, and information specifying the purpose; and a collecting unit that performs a search based on the purpose and keywords and collects the second agent information.

4. The information processing apparatus according to claim 1, wherein the dialogue control unit obtains a first evaluation result for the first AI agent that reflects the consciousness load of the second AI agent and / or a second evaluation result for the second AI agent that reflects the consciousness load of the first AI agent.

5. The information processing apparatus according to claim 4, wherein the dialogue control unit acquires an evaluation result that reflects a value corresponding to the mental load and attentional resources that each AI agent is expected to incur when processing information related to the dialogue.

6. The information processing device according to claim 4, wherein the fewer resources each AI agent references, the lower the evaluation result in terms of cognitive load.

7. The information processing apparatus according to claim 4, wherein the dialogue control unit obtains question information to be asked to the first AI agent from the second AI agent, inputs the question information to the first AI agent to obtain answer information, inputs the question information and the answer information to the second AI agent to obtain an evaluation result for the first AI agent, and the fewer the number of dialogues until the answer information is obtained, the lower the evaluation result with a lower cognitive load is obtained.

8. The information processing apparatus according to claim 4, further comprising a receiving unit that receives a dialogue environment setting including at least one of a dialogue date, a dialogue start time, a dialogue language, a dialogue means, and a dialogue schedule, wherein the dialogue control unit obtains an evaluation result that reflects the conscious load corresponding to the dialogue environment setting.

9. The information processing apparatus according to claim 4, wherein the dialogue control unit obtains question information to be asked to the first AI agent from the second AI agent, inputs the question information to the first AI agent to obtain answer information, inputs the question information and the answer information to the second AI agent to obtain an evaluation result for the first AI agent, and the closer the number of characters in the question information and the number of characters in the answer information are, the lower the evaluation result indicating a lower cognitive load is obtained.

10. The information processing apparatus according to claim 1, wherein the agent creation unit creates a plurality of second AI agents, and the dialogue control unit causes the first AI agent and each of the second AI agents to interact, and obtains the first evaluation result of each of the second AI agents with respect to the first AI agent and / or the second evaluation result of the first AI agent with respect to each of the second AI agents.

11. The information processing apparatus according to claim 10, further comprising a receiving unit that receives information specifying the first agent information, information specifying the second agent information, and information specifying the purpose, wherein the information specifying the second agent information includes conditions, and the agent creation unit creates the plurality of second AI agents based on the conditions.

12. The information processing apparatus according to claim 1, wherein the agent creation unit creates a plurality of first AI agents and a plurality of second AI agents, and the dialogue control unit causes each first AI agent and each second AI agent to interact, and obtains each first evaluation result of each second AI agent with respect to each first AI agent and / or each second evaluation result of each first AI agent with respect to each second AI agent.

13. The information processing apparatus according to claim 10 or 11, wherein the result creation unit creates the matching result based on a first evaluation result and / or a second evaluation result that is higher than a predetermined standard.

14. The information processing apparatus according to claim 2, further comprising a recommendation method storage unit that stores a second behavioral modification method that is recommended to be included in the dialogue content, wherein the result creation unit creates a matching result that includes difference information between the first behavioral modification method and the second behavioral modification method.

15. A method comprising: a computer creating an AI agent based on agent information; causing a first AI agent created based on first agent information and a second AI agent created based on second agent information to interact for a predetermined purpose, thereby obtaining a first evaluation result of the second AI agent with respect to the first AI agent and / or a second evaluation result of the first AI agent with respect to the second AI agent; creating a matching result based on the first evaluation result and / or the second evaluation result; and outputting the matching result.

16. A computer-readable recording medium that records a program causing a computer to perform the following processes: creating an AI agent based on agent information; causing a first AI agent created based on first agent information and a second AI agent created based on second agent information to interact for a predetermined purpose, thereby obtaining a first evaluation result of the second AI agent for the first AI agent and / or a second evaluation result of the first AI agent for the second AI agent; creating a matching result based on the first evaluation result and / or the second evaluation result; and outputting the matching result.

Images