Information processing device

Abstract

To provide highly accurate dialogue. [Solution] The information processing device acquires an utterance made by the user, and based on the utterance, determines and transfers the utterance to one or more agents from among a plurality of agents that are capable of natural language dialogue and are specialized in a plurality of predetermined domains, obtains one or more responses from the one or more agents to which the utterance was transferred, integrates the one or more responses, and provides them to the user.

Description

Technical Field

[0001] This disclosure relates to dialogue technology.

Background Art

[0002] There is a technology for conducting natural language dialogue based on an input text. In this regard, for example, Patent Document 1 discloses an apparatus capable of interacting with a plurality of agents via a network.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] In the future, with the development of machine learning, it is considered that natural language dialogue services will increase further.

[0005] This disclosure aims to provide highly accurate dialogue.

Means for Solving the Problems

[0006] One aspect of an embodiment of this disclosure is obtaining an utterance made by a user, determining, based on the utterance, one or more external agents that can conduct natural language dialogue and to which the utterance is to be transferred among a plurality of external agents specialized for respective plural predetermined domains, transferring the utterance to the determined one or more external agents, obtaining one or more responses from each external agent, integrating the one or more responses, and providing the integrated responses to the user, and an information processing apparatus having a control unit that executes these operations.

[0007] Other embodiments include a method performed by the above-mentioned device, a program for causing a computer to perform the method, or a computer-readable storage medium that non-temporarily stores the program. [Effects of the Invention]

[0008] According to this disclosure, it is possible to provide highly accurate dialogue. [Brief explanation of the drawing]

[0009] [Figure 1] A schematic diagram of the dialogue system according to the first embodiment. [Figure 2] Hardware configuration diagram of the in-vehicle device 10. [Figure 3] Software configuration diagram of the in-vehicle device 10. [Figure 4] This diagram shows the processing flow executed by the control unit 11 of the in-vehicle device 10. [Figure 5] An example of prompt text and response. [Figure 6] A flowchart of the process executed by the control unit 11 of the in-vehicle device 10. [Modes for carrying out the invention]

[0010] In recent years, with the advancement of machine learning, the number of products incorporating language models has increased. For example, by incorporating a Large Language Model (LLM), whose accuracy has been improved through machine learning using large datasets, into a product, the target product can perform natural language processing. Dialogue functionality can be added.

[0011] Automobiles are a prime example of the usefulness of natural language interaction. For instance, by equipping in-car systems with LLM (Language Learning Module), information can be obtained without operating touch panels or other similar devices. Such functionality is particularly useful when the driver's hands are occupied, such as while driving.

[0012] On the other hand, pre-trained language models (LLMs) are large in size and cost-related, making them difficult to implement in embedded devices. Therefore, methods for accessing LLMs via a network have been devised. Accessing LLMs over a network makes it possible to utilize large-scale language models that cannot be stored on local storage.

[0013] Furthermore, by using a network, it becomes possible to selectively access LLMs specialized in specific fields. For example, if multiple LLMs are available, such as an LLM specializing in route guidance, an LLM specializing in local information, and an LLM specializing in casual conversation, it becomes possible to select an LLM that matches the user's intent. In this regard, a technology is known that determines the category of the utterance based on the content of the user's utterance and automatically selects an LLM that matches that category.

[0014] However, if you choose an LLM that focuses on dialogue from among several available LLMs, you may encounter the problem of being limited to dialogues specific to a particular field. For example, if a user makes an utterance that includes something like, "The car's warning light came on," and an LLM (Life Assistance Machine) specializing in vehicle operation guidance is selected, it is expected that the LLM will respond to the utterance by providing information about the meaning of the warning light. However, because an LLM specializing in vehicle operation guidance does not have local information, it may not be able to provide services such as "guiding you to the nearest car dealer for inspection."

[0015] One possible solution to this problem is to transfer the user's utterances to multiple LLMs and then integrate their content. The information processing device relating to this disclosure solves such problems.

[0016] The information processing apparatus according to the first aspect of the present disclosure acquires the utterance made by the user, determines one or more external agents to which the utterance is to be transferred among a plurality of external agents capable of natural language dialogue and specialized for respective plural predetermined domains, transfers the utterance to the determined one or more external agents, acquires one or more answers from each external agent, integrates the one or more answers, and provides the integrated answers to the user, and includes a control unit that executes these operations.

[0017] Each of the plurality of external agents is a language model capable of natural language dialogue and may include a language model specialized for intention understanding in a plurality of domains. The language model may be a large language model (LLM). The large language model is, for example, a language model trained so as to be able to execute a natural language dialogue task. Examples of the plurality of domains include "route guidance", "local information guidance", "vehicle handling guidance", "tourist guidance", and "casual conversation". These agents typically have a configuration that allows remote access via a network.

[0018] The control unit determines one or more external agents to which the utterance made by the user is to be transferred and acquires an answer from the external agent. The control unit may transmit, via the network, prompt text including the content (utterance sentence) of the utterance made by the user to the determined external agent. Thereby, one or more answers can be acquired from each external agent.

[0019] In addition, the control unit integrates one or more answers acquired from each external agent and provides the integrated answers to the user. The integration of the answers may be performed based on a rule or may be performed using another language model. For example, the answers obtained from each external agent may be input to an agent (local agent) that operates locally and has a language model to perform the integration of the answers. The language model possessed by the local agent can be a relatively low-cost language model that can execute natural language dialogue tasks but has not been specialized for a specific domain.

[0020] According to this configuration, it becomes possible to obtain information from multiple LLMs with various specialties and integrate them to provide to the user. In the case of the example described above, for example, by integrating the answer obtained from the LLM specialized in vehicle handling guidance and the answer obtained from the LLM specialized in local information guidance, it becomes possible to "guide the meaning of warning lights and also guide service bases that can solve problems".

[0021] Further, the control unit may further obtain context information corresponding to the user and further input the obtained context information to one or more external agents that transfer the utterance. Examples of context information corresponding to the user can include, for example, user attribute information and information related to the movement of the user. The user attribute information may include, for example, the user's age, gender, preferences, etc. The information related to the movement of the user may include, for example, the destination and route information of the vehicle in which the user is riding (the vehicle equipped with the information processing apparatus according to the present disclosure).

[0022] Hereinafter, specific embodiments of the present disclosure will be described based on the drawings. The hardware configuration, module configuration, functional configuration, etc. described in each embodiment are not intended to limit the technical scope of the disclosure only to those unless otherwise specified.

[0023] (First Embodiment) [Overview of the System] An overview of the dialogue system according to the first embodiment will be described. The dialogue system according to this embodiment is configured to include an in-vehicle device 10 mounted on a vehicle. The in-vehicle device 10 can access an agent (external agent) that provides natural language dialogue services via a network (e.g., a mobile communication network).

[0024] The in-vehicle device 10 is installed in a connected vehicle that can communicate with any device via wireless communication. The in-vehicle device 10 may include a data communication module (DCM) for connecting vehicle components (e.g., ECUs and in-vehicle terminals) to a network. In this embodiment, the in-vehicle device 10 can access the internet via a predetermined mobile communication network and connect to an agent that provides conversational services.

[0025] Furthermore, the vehicle (or components installed in the vehicle) can provide various services by communicating with external devices via the in-vehicle device 10. Examples of such services include navigation services, remote control services (e.g., remote air conditioning), in-vehicle Wi-Fi® services, and emergency call services.

[0026] Furthermore, the in-vehicle device 10 has an audio input / output function, and communicates with the vehicle occupants in natural language. Dialogue can be conducted.

[0027] Figure 1(A) is a schematic diagram illustrating a conventional in-vehicle device 10. The in-vehicle device 10 comprises a control unit for voice input / output and voice recognition, and a local agent. The local agent includes a language model trained to enable natural language dialogue. The local agent may provide predetermined functions, such as route planning or vehicle operation (air conditioning, etc.), through dialogue. In the illustrated configuration, the language model used by the local agent is stored in the memory device of the in-vehicle device 10.

[0028] On the other hand, such a configuration makes it difficult to provide conversational services using large-scale language models. For example, conversational categories are diverse, including local information, tourist information, vehicle information (operating manuals), and casual conversation. To cover all of these categories, it is necessary to prepare a language model trained using a large amount of training data. Such language models can range in size from several gigabytes to several hundred gigabytes, and storing them in local storage is not practical from a cost perspective.

[0029] Therefore, systems have been proposed that allow access to language models specialized for dialogue within specific domains, via a network. Figure 1(B) is a schematic diagram of a configuration in which the in-vehicle device 10 can access not only local agents but also multiple language models provided on the internet. For example, one possible configuration is to place agents (external agents) that have multiple large-scale language models specialized for a specific domain on a device connected to the internet, and access them from the in-vehicle device 10 as needed.

[0030] However, this model has a problem: it limits conversations to a specific domain. For example, if a user wants to obtain information about a particular sport, they would likely interact with an agent specializing in that sport. However, since this agent cannot provide information unrelated to the sport, such as facility information, it cannot, for instance, guide the user to a venue where the sport can be watched.

[0031] Therefore, in the in-vehicle device 10 according to this embodiment, the local agent is provided with a function to integrate responses from multiple external agents. User utterances are forwarded to multiple external agents, and responses are obtained from these multiple external agents. These responses are integrated in the local agent, and the integrated response is provided to the user. Figure 1(C) is a schematic diagram of the in-vehicle device 10 according to this embodiment.

[0032] When the in-vehicle device 10 recognizes a user's utterance, it determines, based on the content of the utterance, which external agent is appropriate for that utterance. This process may be performed using a rule-based method or a machine learning model. The control unit then forwards the utterance to the determined external agent.

[0033] The responses from the selected external agents are integrated by the local agent and provided to the user. This will enable the appropriate provision of responses obtained from multiple external agents to the user.

[0034] In this embodiment, the terms "access to an external agent" and "connection to an external agent" refer to access to or connection to an external device that provides dialogue services by an external agent. In this embodiment, each of the multiple external devices is configured as follows: Multiple dialogue services based on large-scale language models operate as "external agents," and the in-vehicle device 10 can engage in dialogue by connecting to any of these external agents.

[0035] [Hardware configuration] Next, the hardware configuration of each device constituting the system will be described. Figure 2 is a schematic diagram showing an example of the hardware configuration of the in-vehicle device 10 according to this embodiment.

[0036] The in-vehicle device 10 can be configured as a computer having a processor (CPU, GPU, etc.), main memory (RAM, ROM, etc.), and auxiliary storage (EPROM, hard disk drive, removable media, etc.). The auxiliary storage contains an operating system (OS), various programs, various tables, etc., and by executing the programs stored therein, various functions (software modules) that match a predetermined purpose, as described later, can be realized. However, some or all of the functions may be realized as hardware modules by hardware circuits such as ASICs and FPGAs.

[0037] The in-vehicle device 10 is configured as hardware, comprising a control unit 11, a storage unit 12, a wireless communication module 13, and an input / output unit 14.

[0038] The control unit 11 is a computing unit that realizes various functions of the in-vehicle device 10 by executing a predetermined program. The control unit 11 can be implemented by a hardware processor such as a CPU. The control unit 11 may also be configured to include RAM, ROM (Read Only Memory), cache memory, etc.

[0039] The memory unit 12 is a means for storing information and is composed of storage media such as RAM, magnetic disks, and flash memory. The memory unit 12 stores programs executed by the control unit 11, data used by those programs, and so on.

[0040] The wireless communication module 13 is a communication device that performs wireless communication with a predetermined network. In this embodiment, the wireless communication module 13 is configured to communicate with a predetermined mobile communication network. The wireless communication module 13 can be configured to have an eUICC (e.g., a SIM card). The SIM card is configured as a microcomputer equipped with a CPU and a memory device. The SIM card stores information for connecting to the mobile communication network and receiving authentication.

[0041] The input / output unit 14 is a unit that receives input from the user of the device and presents information to the user. Typically, the input / output unit 14 includes devices for inputting and outputting sound, such as a microphone and a speaker. The input / output unit 14 may also include devices that provide visual information (such as a display).

[0042] [Software Configuration] Next, the software configuration of each device constituting the system will be described. Figure 3 is a schematic diagram showing the software configuration of the in-vehicle device 10 according to this embodiment.

[0043] In this embodiment, the control unit 11 of the in-vehicle device 10 is configured to have three software modules: a dialogue reception unit 111, an agent unit 112, and a route guidance unit 113. Each software module may be implemented by the control unit 11 (CPU, etc.) executing a program stored in the storage unit 12. The information processing performed is synonymous with the information processing performed by the control unit 11 (CPU, etc.).

[0044] The dialogue reception unit 111 acquires utterances made by the vehicle occupants (hereinafter also referred to as users) via the input / output unit 14. The dialogue reception unit 111 performs predetermined processing on the acquired voice data and performs speech recognition. This converts the content of the utterances into text. The dialogue reception unit 111 also transmits the text obtained as a result of speech recognition to the agent unit 112.

[0045] Furthermore, the dialogue reception unit 111 outputs the response (hereinafter referred to as the answer text) from the language model transmitted from the agent unit 112. The dialogue reception unit 111 converts the answer text into speech and outputs it via the input / output unit 14.

[0046] The agent unit 112 first provides a dialogue service using a language model stored in its own device (local language model 12A, described later). The agent unit 112 can behave as a virtual agent (local agent). The local agent can, for example, provide a dialogue service that is completed within its own device (e.g., casual conversation).

[0047] Furthermore, when specialized knowledge is required in the dialogue, the agent unit 112 forwards the utterance to one or more external agents available via the network and retrieves the results. For example, if it is determined that the user is seeking information about a vehicle, the agent unit 112 selects an external agent capable of providing vehicle operation guidance. Also, if it is determined that the user is seeking information about a facility or store, the agent unit 112 selects an external agent capable of providing local information.

[0048] The selection of an external agent may be based on the results of analyzing the content of the utterance. For example, the agent unit 112 may determine the degree of suitability for each of several external agents based on the words and context contained in the utterance, and transfer the utterance to the external agent that has obtained a degree of suitability equal to or greater than a predetermined value.

[0049] Furthermore, the agent unit 112 may select multiple external agents to which the utterance is forwarded. For example, if there are multiple external agents whose relevance to the utterance is above a predetermined value, the agent unit 112 may forward the utterance to these multiple external agents. In this case, the agent unit 112 obtains the responses to the utterance from the multiple external agents, integrates these responses, and provides them to the user.

[0050] The route guidance unit 113 receives the user's specified departure point and destination, and generates a route connecting the departure point and destination based on road map data, etc. It also outputs information about the generated route via the input / output unit 204.

[0051] The storage unit 12 of the in-vehicle device 10 stores a local language model 12A, agent information 12B, and user information 12C.

[0052] Local Language Model 12A is a language model trained to perform natural language dialogue tasks. Local Language Model 12A is a relatively low-cost language model and is not trained for domain-specific dialogue. Any language model capable of general-purpose conversation can be used as Local Language Model 12A. For example, an open-source language model can be used as Local Language Model 12A.

[0053] The local language model 12A is a lighter language model than the large-scale language models possessed by external agents accessible via the network. By utilizing the local language model 12A, the agent unit 112 can engage in responsive conversations. For example, it becomes possible to transfer utterances to an external agent when specialized conversations are required, and have the local agent process utterances when conversations requiring fast responses are needed.

[0054] Agent information 12B is a collection of data relating to multiple external agents available to the in-vehicle device 10. Agent information 12B may include, for example, the identifier and name of the external agent, information about the domain it specializes in, and information about the access destination (for example, the network address of the external device providing the external agent). The agent unit 112 can determine the destination of the utterance by referring to agent information 12B.

[0055] User information 12C is a collection of data relating to the attributes of the user interacting with the in-vehicle device 10. User information 12C may include, for example, information about the user's preferences, as well as personal information such as the user's gender, age, height, and weight. If there are multiple users of the vehicle, user information 12C may also include data for multiple people.

[0056] [Flow of the conversation] Next, we will explain the overview of the processing performed by the control unit 11. Figure 4 is a diagram showing the flow of processing performed by the control unit 11 of the in-vehicle device 10 when it receives a speech from the user.

[0057] The dialogue reception unit 111 acquires utterances from the user via the input / output unit 14. For example, the input / output unit 14 converts utterances acquired via a microphone or the like into voice data, which the dialogue reception unit 111 acquires. The dialogue reception unit 111 performs a predetermined speech recognition process on the acquired voice data and converts the voice data into text. While holding off on responding to the utterance, the dialogue reception unit 111 transmits the text obtained as a result of the speech recognition to the agent unit 112. This text will hereafter be referred to as the "utterance text".

[0058] Upon receiving the utterance, the agent unit 112 determines the intent of the utterance (what the user wants) and the user's situation (what they are currently doing) based on the content of the utterance, and then determines the external agent to which the utterance will be forwarded. The intent of the utterance may be determined using the local language model 12A, or using other machine learning models. It may also be determined using a rule-based method. Furthermore, the user's status may be determined by using sensor data other than speech. For example, the agent unit 112 may determine the user's status by using data such as vehicle speed, current position, and direction of travel obtained from sensors on the vehicle. Examples of user statuses include "traveling on a regular road," "traveling on a highway," "stopped," "waiting at a traffic light," "traveling to work," and "going home."

[0059] The agent unit 112 obtains information about external agents (agent information 12B) from the storage unit 12 and determines which external agent to transfer the utterance to based on the agent information 12B and the determined intent and circumstances. The agent unit 112 may, for example, determine the degree of suitability for each of multiple external agents and decide to transfer the utterance to the external agent that has obtained a degree of suitability equal to or greater than a predetermined value.

[0060] In this example, we assume that the following four external agents are available. (1) Vehicle Agent An external agent capable of monitoring the vehicle's interior and configuring in-vehicle equipment. It has learned the vehicle's owner's manual and can also provide information about the vehicle's functions. (2) Casual Conversation Agent An external agent specializing in non-task-oriented conversation. (3) Regional information agents An external agent capable of providing local area information, including store and tourist information. In this example, a vehicle agent and a local information agent are selected as external agents that are suitable for the utterance.

[0061] The agent unit 112 may also determine that it is not necessary to forward the utterance to an external agent. For example, this applies when the user's request is for route guidance and can be completed within the device itself, or when a fast response is required. In this case, the utterance is processed only by the agent unit 112 (and the local language model 12A).

[0062] Next, the agent unit 112 generates prompt text to be input to the selected external agent. Figure 5(A) shows an example of prompt text input from the agent unit 112 to the external agent. The prompt text includes not only the spoken utterance but also contextual information relevant to the user. In this embodiment, context information refers to user attribute information and information about the user's movements. Context information may include, for example, information about the user's preferences, or it may include personal information such as the user's gender, age, height, and weight. This information may be obtained from user information 12C stored in the storage unit 12. Furthermore, contextual information may include information about the user's movement. Examples of such information include information about the vehicle's departure point, destination, travel route, and intermediate stops. This information may be obtained from the route guidance unit 113.

[0063] The agent unit 112 transfers the generated prompt text to an external device providing the desired external agent via the network.

[0064] An external agent that receives a spoken sentence outputs a response in response to that sentence. Figure 5(B) shows an example of a response from a vehicle agent, and Figure 5(C) shows an example of a response from a local information agent. As illustrated, each agent outputs a response related to the domain in which it specializes. For example, the vehicle agent outputs information about the specific meaning of warning lights, and the local information agent outputs information about spots where vehicle inspections can be performed.

[0065] When the local language model 12A receives a response from the external agent that transmitted the utterance, it integrates the response and sends the result as a reply to the dialogue reception unit 111. Figure 5(D) shows an example of a reply obtained as a result of the integration. Response integration can be performed using the local language model 12A. For example, the agent unit 112 inputs a prompt text containing one or more received responses to the local language model 12A, causing it to perform response integration. This prompt text includes, for example, an instruction to integrate the contents of the responses obtained from each external agent and generate a new sentence. As a result, the local language model 12A outputs the integrated response sentence. The integrated response sentence is then sent from the agent unit 112 to the dialogue reception unit 111.

[0066] When the dialogue reception unit 111 receives a response, it generates audio data based on that response. The output is sent via the input / output unit 14. The dialogue reception unit 111 may generate only audio data that reads out the response text, or it may generate information associated with the audio data. An example of information associated with the audio data is a user interface screen. The user interface screen may include the response text written in text. The dialogue system according to this embodiment seamlessly performs the series of processes shown in Figure 4. As a result, from the perspective of the user who made the utterance, it is possible to obtain the effect of interacting with the in-vehicle device 10.

[0067] [flowchart] Next, the details of the processes performed by the in-vehicle device 10 will be described. Figure 6 is a flowchart of the processes performed by the in-vehicle device 10. The processes shown in Figure 6 are initiated when a user (vehicle occupant) makes a speech. The start of a speech may be detected, for example, by a predetermined keyword.

[0068] First, in step S11, the dialogue reception unit 111 recognizes the content of the utterance. The dialogue reception unit 111 acquires the audio data output from the input / output unit 14 and performs speech recognition processing to convert the utterance into text. The converted text (utterance) is sent to the agent unit 112.

[0069] In step S12, the agent unit 112 determines an external agent to which the utterance and context information will be transferred. The external agent to which the utterance and context information will be transferred may, for example, be determined by estimating the user's intent (what the user is requesting) and situation (what the user is currently doing) from the utterance. For example, the agent unit 112 may read information about the characteristics of an external agent (agent information 12B) from the storage unit 12 and determine which external agent to use based on that information.

[0070] Next, in step S13, the agent unit 112 acquires the user's context information. If the context information is related to the user's movement, it may be acquired from the route guidance unit 113. Alternatively, if the context information is related to the user's attributes, it may be acquired from the storage unit 12.

[0071] In step S14, the agent unit 112 transfers the utterance and context information to the external agent determined in step S12. For example, the agent unit 112 accesses an external device providing a service by the target external agent via the wireless communication module 13 and transfers the utterance and context information to that external agent. In step S15, the agent unit 112 obtains a response from the target external agent.

[0072] In step S16, the agent unit 112 integrates these responses using the local language model 12A. Specifically, the agent unit 112 generates a list of responses obtained from multiple external agents and prompt text containing instructions to integrate these responses, and inputs this prompt text into the local language model 12A. It also retrieves the integrated response text output from the local language model 12A.

[0073] In step S17, the agent unit 112 transmits the response sentence obtained as a result of the integration to the dialogue reception unit 111. The dialogue reception unit 111 generates voice data based on the response sentence and outputs it via the input / output unit 14. The dialogue reception unit 111 uses speech synthesis technology to convert the response sentence into voice. The audio data may be converted into voice data. The voice data is output via the input / output unit 14 (speaker, etc.) and provided to the user.

[0074] As described above, in this embodiment, the in-vehicle device 10 has a local agent that determines which external agent should forward the utterance made by the vehicle occupant. The local agent also integrates and outputs the response sentences obtained from the external agents. This makes it possible to provide the user with a single response obtained from multiple external agents that handle highly specialized topics.

[0075] (modified version) The embodiments described above are merely examples, and this disclosure may be modified as appropriate without departing from its essence. For example, the processes and means described in this disclosure can be freely combined and implemented, as long as no technical inconsistencies arise.

[0076] Furthermore, while the embodiment illustrates a configuration in which the external agent is deployed on an external device connected to the Internet, the external agent may be deployed on other devices. For example, the external agent may be deployed on an edge server accessible from a vehicle. Alternatively, the external agent may be distributed across multiple edge servers or the cloud. Furthermore, although the embodiments illustrate in-vehicle devices, the information processing device according to this disclosure can also be implemented as a device or equipment that is not installed in a vehicle.

[0077] Furthermore, if external agents are deployed on multiple edge servers, the in-vehicle device 10 may acquire the vehicle's location information and identify the edge server where the desired external agent is located based on that location information. For example, an external agent providing regional information may be deployed on multiple edge servers, each deployed for a specific region. In this case, for example, the in-vehicle device 10 may identify the external agent deployed on the geographically closest edge server based on the vehicle's location information. The regional information handled by the external agent may differ from one edge server to another. For example, an edge server deployed in area A may have an external agent providing regional information for area A, and an edge server deployed in area B may have an external agent providing regional information for area B. The in-vehicle device 10 can identify the edge server corresponding to the area where the vehicle is located and obtain the regional information for that area by accessing the external agent provided by that edge server.

[0078] Furthermore, a process described as being performed by a single device may be divided and executed by multiple devices. Conversely, a process described as being performed by different devices may be executed by a single device. In a computer system, the hardware configuration (server configuration) by which each function is implemented can be flexibly changed.

[0079] This disclosure can also be realized by supplying a computer program implementing the functions described in the embodiments above to a computer, and having one or more processors in the computer read and execute the program. Such a computer program may be provided to the computer by a non-temporary computer-readable storage medium that can be connected to the computer's system bus, or it may be provided to the computer via a network. The non-temporary computer-readable storage medium may be any type of disk, such as magnetic disks (floppy disks, hard disk drives (HDDs), etc.), optical disks (CD-ROMs, DVDs, Blu-ray discs, etc.), read-only memory (ROM), random access memory (RAM), EPROM, EEPROM, magnetic cards, flash memory, optical cards, or any other medium suitable for storing electronic instructions. Includes the medium of the ipu. [Explanation of Symbols]

[0080] 10...In-vehicle equipment 11. Control Unit 12...Storage section 13. Wireless communication module 14...Input / output section

Claims

1. To obtain the utterances made by the user, Based on the aforementioned utterance, the system determines one or more external agents to forward the utterance from among multiple external agents, each specialized in a predetermined domain and capable of natural language dialogue. The utterance is forwarded to one or more external agents determined in the preceding steps, and one or more responses are obtained from each external agent. To integrate one or more of the above responses and provide them to the user, An information processing device having a control unit that performs the following.

2. The control unit further acquires context information corresponding to the user and further transmits the acquired context information to one or more external agents that forward the utterance. The information processing apparatus according to claim 1.

3. The context information includes the user's attribute information or information regarding the user's movement. The information processing apparatus according to claim 2.

4. The context information includes destination or route information of the vehicle on which the information processing device is installed. The information processing apparatus according to claim 2.

5. The control unit inputs the one or more responses to a local agent capable of natural language dialogue, thereby integrating the one or more responses. The information processing apparatus according to claim 1.

Images