Systems, methods, and non-transitory computer readable media

Abstract

The present disclosure provides systems, methods, and non-transitory computer-readable media. A technique for improving an AI chat robot including a plurality of dialogue-type models. A system (1) is provided with a communication device (10) and a server (20). A server (20) stores a first dialogue-type model and a second dialogue-type model. Furthermore, the communication device (10) displays a first screen (40) on which a conversation between the user and the first conversation-type model is performed, and transmits the question input by the user to the first screen (40) to the server. The server (20) determines whether or not a predicted time required until the first dialogue-type model generates an answer to the question content related to the question is longer than a predetermined reference. Furthermore, when the server (20) determines that the predicted time is longer than the predetermined reference time, the server (20) executes a conversation between the user and the second dialogue-type model until the completion of the generation of the answer by the first dialogue-type model.

Description

Technical Field

[0001] This disclosure relates to systems, methods, and procedures. Background Technology

[0002] Previously, technologies for AI chatbots that include multiple conversational models were known. For example, Patent Document 1 discloses a technology for an AI chatbot that includes multiple conversational models, which selects a conversational model based on the user's question and engages in conversation with the user.

[0003] Existing technical documents

[0004] Patent documents

[0005] Patent Document 1: Japanese Patent Application Publication No. 2020-091513 Summary of the Invention

[0006] There is room for improvement in the technology of AI chatbots that include multiple conversational models.

[0007] The purpose of this disclosure, made in light of the above circumstances, is to improve the technology of AI chatbots that include multiple conversational models.

[0008] One embodiment of this disclosure relates to a system equipped with a communication device and a server, wherein...

[0009] The server stores the first dialogic model and the second dialogic model.

[0010] The communication device:

[0011] The first screen displays the dialogue between the user and the first conversational model.

[0012] The question text entered by the user on the first screen is sent to the server.

[0013] The server:

[0014] Determine whether the prediction time required for the first conversational model to generate an answer to the question related to the question text is longer than a predetermined benchmark.

[0015] If it is determined that the predicted time is longer than the predetermined benchmark, the dialogue between the user and the second conversational model is executed until the generation of the response through the first conversational model is completed.

[0016] One embodiment of this disclosure relates to a method executed by a system having a communication device and a server, wherein the method includes:

[0017] The server stores the first dialog model and the second dialog model;

[0018] The communication device:

[0019] The first screen displays the dialogue between the user and the first conversational model;

[0020] The question text entered by the user on the first screen is sent to the server;

[0021] The server:

[0022] Determine whether the prediction time required for the first conversational model to generate an answer to the question content related to the question text is longer than a predetermined benchmark; and

[0023] If it is determined that the predicted time is longer than the predetermined benchmark, the dialogue between the user and the second conversational model is executed until the generation of the response through the first conversational model is completed.

[0024] One embodiment of this disclosure relates to a program that causes a system equipped with a communication device and a server to perform actions including the following steps:

[0025] The server stores the first dialog model and the second dialog model;

[0026] The communication device:

[0027] The first screen displays the dialogue between the user and the first conversational model;

[0028] The question text entered by the user on the first screen is sent to the server;

[0029] The server:

[0030] Determine whether the prediction time required for the first conversational model to generate an answer to the question content related to the question text is longer than a predetermined benchmark; and

[0031] If it is determined that the predicted time is longer than the predetermined benchmark, the dialogue between the user and the second conversational model is executed until the generation of the response through the first conversational model is completed.

[0032] According to one embodiment of this disclosure, the technology for AI chatbots comprising multiple conversational models is improved. Attached Figure Description

[0033] Figure 1 This is a block diagram illustrating the general structure of a system according to one embodiment of this disclosure.

[0034] Figure 2 This is a flowchart illustrating the actions of the server.

[0035] Figure 3 This is a flowchart illustrating the operation of a communication device.

[0036] Figure 4 This is a diagram showing an example of a dialogue screen with the first dialogue model.

[0037] Figure 5 This is a diagram showing an example of displaying a dialogue screen of the second dialogue model overlaid on the dialogue screen of the first dialogue model.

[0038] Figure 6 This is an example of displaying the dialogue and questions from the second conversational model on the screen showing the dialogue with the first conversational model.

[0039] Figure 7 This is a diagram showing an example of being able to distinguish between dialogue with the first dialogic model and dialogue with the second dialogic model. Detailed Implementation

[0040] The following describes the implementation of this disclosure.

[0041] (Summary of the implementation method)

[0042] Reference Figure 1 This section outlines the system 1 involved in the embodiments of this disclosure. System 1 includes a communication device 10 and a server 20. A user can operate the communication device 10 to engage in dialogue with an AI chatbot stored on the server. If the server 20 determines that it will take time to generate an answer to the user's question, it will engage in dialogue with the user using an AI chatbot different from the one generating the answer, until the answer is generated. The communication device 10 and the server 20 are communicatively connected to a network 30, such as the Internet and mobile communication networks.

[0043] The communication device 10 may be, for example, a computer, a smartphone, or a navigation device mounted in a vehicle, but is not limited to these, and may be any communication device. The number of communication devices 10 provided in system 1 may also be arbitrarily determined.

[0044] Server 20 is one or more computers that can communicate with each other and communicate with communication device 10 via network 30.

[0045] First, an overview of this embodiment will be given, with details to follow. System 1 includes a communication device 10 and a server 20. Server 20 stores a first dialog model and a second dialog model. Furthermore, communication device 10 displays a first screen 40 for dialogue between the user and the first dialog model, and sends a question input by the user to the first screen 40 to the server. Server 20 determines whether the prediction time required for the first dialog model to generate an answer related to the question is longer than a predetermined benchmark. If server 20 determines that the prediction time is longer than the predetermined benchmark, server 20 executes the dialogue between the user and the second dialog model until the answer from the first dialog model is generated.

[0046] Thus, according to this embodiment, if it is determined that the predicted time required to generate an answer to a user's question is longer than a predetermined baseline, the server 20 executes a dialogue between the user and the second conversational model until the generation of an answer through the first conversational model is completed. This reduces the perceived time for the user to generate an answer through the first conversational model. Furthermore, by engaging in dialogue with the second conversational model, the user can recognize that the system 1 is functioning correctly. Therefore, the technology of an AI chatbot comprising multiple conversational models is improved in terms of both reducing the perceived time for the user and recognizing that the system 1 is functioning correctly.

[0047] Next, the structure of System 1 will be described in detail.

[0048] (Structure of communication device 10)

[0049] like Figure 1 As shown, the communication device 10 includes a communication unit 11, an output unit 12, an input unit 13, a control unit 14, and a storage unit 15.

[0050] The communication unit 11 includes one or more communication interfaces connected to the network 30. These communication interfaces may correspond to mobile communication standards such as 4G (4th generation) or 5G (5th generation), but are not limited to these. In this embodiment, the communication device 10 communicates with the server 20 via the communication unit 11 and the network 30.

[0051] The output unit 12 includes one or more output devices for outputting information. These output devices may be, for example, a display or a speaker, but are not limited to these. Alternatively, the output unit 12 may also include an interface for connecting an external output device.

[0052] The input unit 13 includes one or more input devices for detecting user input. These input devices may be, for example, physical keys, capacitive keys, pointing devices such as a mouse, a touchscreen integrated with the display of the output unit 12, or a microphone for receiving sound input, but are not limited to these. Alternatively, the input unit 13 may also include an input interface that detects user input via an external input device.

[0053] The control unit 14 includes one or more processors, one or more programmable circuits, one or more dedicated circuits, or combinations thereof. The processor may be a general-purpose processor such as a CPU (Central Processing Unit) or a GPU (Graphics Processing Unit), or a dedicated processor for specific processing, but is not limited to these. The programmable circuit may be, for example, a FPGA (Field-Programmable Gate Array), but is not limited to these. The dedicated circuit may be, for example, an ASIC (Application Specific Integrated Circuit), but is not limited to these. The control unit 14 controls the overall operation of the communication device 10.

[0054] Storage unit 15 includes one or more memory units. These memory units may be, for example, semiconductor memory, magnetic memory, or optical memory, but are not limited to these. Each memory unit included in storage unit 15 may function as a main storage device, an auxiliary storage device, or a cache memory. Storage unit 15 stores any information used in the operation of communication device 10. For example, storage unit 15 may also store system programs, application programs, and embedded software. The information stored in storage unit 15 may be updated using information obtained from network 30 via communication unit 11.

[0055] (Structure of server 20)

[0056] like Figure 1 As shown, the server 20 includes a communication unit 21, an output unit 22, an input unit 23, a control unit 24, and a storage unit 25.

[0057] The communication unit 21 includes one or more communication interfaces connected to the network 30. These communication interfaces may correspond to mobile communication standards such as 4G (4th generation) or 5G (5th generation), but are not limited to these. In this embodiment, the server 20 communicates with the communication device 10 via the communication unit 21 and the network 30.

[0058] The output unit 22 includes one or more output devices for outputting information. These output devices may be, for example, a display or a speaker, but are not limited to these. Alternatively, the output unit 22 may also include an interface for connecting an external output device.

[0059] The input unit 23 includes one or more input devices for detecting user input. These input devices may be, for example, physical keys, capacitive keys, pointing devices such as a mouse, a touchscreen integrated with the display of the output unit 22, or a microphone for receiving sound input, but are not limited to these. Alternatively, the input unit 23 may also include an input interface that detects user input via an external input device.

[0060] The control unit 24 includes one or more processors, one or more programmable circuits, one or more dedicated circuits, or combinations thereof. The processor may be a general-purpose processor such as a CPU (Central Processing Unit) or a GPU (Graphics Processing Unit), or a dedicated processor for specific processing, but is not limited to these. The programmable circuit may be, for example, a FPGA (Field-Programmable Gate Array), but is not limited to these. The dedicated circuit may be, for example, an ASIC (Application Specific Integrated Circuit), but is not limited to these. The control unit 24 controls the overall operation of the server 20.

[0061] Storage unit 25 includes one or more memory units. These memory units may be, for example, semiconductor memory, magnetic memory, or optical memory, but are not limited to these. Each memory unit included in storage unit 25 may function as a main storage device, an auxiliary storage device, or a cache memory. Storage unit 25 stores any information used in the operation of server 20. For example, storage unit 25 may also store system programs, application programs, and embedded software. Additionally, storage unit 25 may store an AI chatbot including a first conversational model and a second conversational model. The first conversational model is a model that generates and outputs answers to user questions. The second conversational model is a model that engages in dialogue with the user until the generation of an answer from the first conversational model is completed. The second conversational model is a model that requires less time to generate dialogue compared to the first conversational model.

[0062] (The action flow of server 20)

[0063] Reference Figure 2 This describes the operation of the server 20 involved in this embodiment.

[0064] S100: The control unit 24 begins a dialogue with the first dialog model stored in the storage unit 25.

[0065] The dialogue with the first dialogue model can also begin when the server 20 receives information from the user that the user has started a dialogue with the first dialogue model by operating the communication device 10, but it is not limited to this and can also determine any starting conditions.

[0066] S101: Input unit 23 receives a query text input by the user into input unit 13 via network 30.

[0067] The question text is the text entered by the user into input section 13. Any input method can be used.

[0068] S102: The control unit 24 determines whether the prediction time required for the first dialogic model to generate an answer to the question content related to the question text is longer than the predetermined benchmark.

[0069] The query content is derived from information extracted from the user-input query. The information extraction method can employ any technique, including query transformations such as natural language processing. For example, if the query content related to the user-input query is included in the historical information (described later), it can be determined that the prediction time is shorter than a predetermined benchmark. Conversely, if the query content related to the user-input query is not included in the historical information, it can be determined that the prediction time is longer than a predetermined benchmark. If the prediction time is determined to be longer than the predetermined benchmark, step S103 is executed; if the prediction time is determined to be shorter than the predetermined benchmark, step S104 is executed.

[0070] S103: Control unit 24 executes the dialogue between the user and the second conversational model until the generation of the response through the first conversational model is completed.

[0071] If the control unit 24 determines that the predicted time is longer than a predetermined reference, it will, during the period until the generation of the response from the first conversational model is completed, execute a dialogue with the second conversational model via the network 30 using the user-operated communication device 10. This reduces the perceived time for the user to generate the response from the first conversational model. Furthermore, by executing the dialogue with the second conversational model, the user can verify that the system 1 is functioning correctly.

[0072] S104: The control unit 24 sends the response (text) generated by the first dialogue model to the communication device 10 via the network 30.

[0073] Control unit 24 may, for example, send a response to communication device 10. Control unit 24 may also send information related to the response and use communication device 10 to generate a response.

[0074] S105: The control unit 24 stores the question content and the answer content in the storage unit 25.

[0075] The storage unit 25 stores historical information including one or more previously input questions and one or more answers to those questions. The control unit 24 can also determine that the prediction time is short if the question content related to the user-input question is included in the historical information of the storage unit 25, and then use the first conversational model to extract the answer content corresponding to the question content from the historical information, and generate an answer to the question content based on the extracted answer content. This reduces the user's waiting time until the answer generated by the first conversational model is completed.

[0076] (Operation flow of communication device 10)

[0077] Reference Figure 3 This explains the operation of the communication device 10 involved in this embodiment.

[0078] S200: The control unit 14 displays the first screen 40 on the output unit 12, which is the first screen for dialogue between the user and the first dialogic model.

[0079] Screen 40 is the screen used for the user to have a dialogue with the first conversational model. Figure 4 This diagram illustrates an example of the first screen 40. The first screen 40 includes a user's question 41, a first conversational model's answer 42, and another user's question 43. The user's question 41 is something like, "What function is disabled when a feature like forward-looking driver assistance is turned off?" The first conversational model's answer 42 is a description of functions related to forward-looking driver assistance, including an animated icon 44. Clicking the animated icon 44 plays an animation. The user's question 43 is something like, "I only want to play the same music. I also want to play other music." The first screen 40 includes, but is not limited to, an input section for the user to enter their question and a display section for providing an answer to the user's question; it can display any information.

[0080] S201: The control unit 14 sends the query text entered by the user into the first screen 40 to the server 20.

[0081] The user enters a question into the input section 13. The entered question is sent to the server 20 via the network 30. The question can be entered using any method, such as text input or voice input.

[0082] S202: Control unit 14 determines whether the dialogue with the second dialogue model has started.

[0083] The control unit 14 determines whether the server 20 has decided to initiate a dialogue between the user and the second conversational model. For example, the control unit 14 may receive information from the server 20 via the network 30 indicating that a dialogue with the second conversational model has begun, but it is not limited to this; any method can be used to determine this. If a dialogue with the second conversational model has begun using the server 20, S203 is executed; if a dialogue with the second conversational model has not begun, S204 is executed.

[0084] S203: The control unit 14 overlays a second screen 50 on the output unit 12, showing the dialogue between the user and the second dialogue model.

[0085] Screen 50 is the screen used for the user to have a dialogue with the second conversational model. Figure 5 An example is shown where a second screen 50 is overlaid on a first screen 40. The second screen 50 includes, but is not limited to, an input section for the user to input a question and a display section for providing an answer to the user's question; it can display any information. Furthermore, the second screen 50 is not limited to overlay display and can employ any display method. Thus, the user can recognize that system 1 is working properly. Additionally, in Figure 5 In the example, the user engages in a dialogue on screen 50 that is unrelated to the question posed to the first conversational model, such as "Do you have any hobbies?", but is not limited to this and can engage in dialogue with any content.

[0086] S204: Control unit 14 receives the response (text) generated by the first conversational model from server 20 via network 30.

[0087] Regarding the response, for example, an article can be received from server 20. Alternatively, information related to the response can be received, and an article can be created using communication device 10.

[0088] S205: The control unit 14 displays the response text on the output unit 12.

[0089] For example, when the control unit 14 receives a response from the server 20, it displays the received response on the output unit 12. Alternatively, when a response is received from the server 20, the control unit 14 can also create a response based on the response and display it on the output unit 12. Furthermore, the output method can employ any method, such as text display or sound output.

[0090] As described above, System 1 includes a communication device 10 and a server 20. Server 20 stores a first dialog model and a second dialog model. Furthermore, the communication device 10 displays a first screen 40 for dialogue between the user and the first dialog model, and sends a question input by the user to the first screen 40 to the server. Server 20 determines whether the prediction time required for the first dialog model to generate an answer related to the question is longer than a predetermined benchmark. If server 20 determines that the prediction time is longer than the predetermined benchmark, server 20 executes the dialogue between the user and the second dialog model until the answer from the first dialog model is generated.

[0091] According to the above structure, if it is determined that the predicted time required to generate an answer to a user's question is longer than a predetermined baseline, the server 20 executes a dialogue between the user and the second conversational model until the generation of an answer through the first conversational model is completed. This reduces the perceived time for the user to generate an answer through the first conversational model. Furthermore, by engaging in dialogue with the second conversational model, the user can recognize that System 1 is functioning correctly. Therefore, the technology of an AI chatbot comprising multiple conversational models is improved in terms of both reducing the perceived time for the user and recognizing that System 1 is functioning correctly.

[0092] This disclosure has been described with reference to the accompanying drawings and embodiments, but it should be noted that those skilled in the art can make various modifications and alterations based on this disclosure. Therefore, it is intended that such modifications and alterations be included within the scope of this disclosure. For example, the functions included in each structural part or step can be logically reconfigured without contradiction, and multiple structural parts or steps can be combined into one or divided.

[0093] For example, in the above embodiments, it is also possible to implement an embodiment in which the structure and operation of the server 20 are distributed among multiple computers capable of communicating with each other. For example, it is also possible to implement an embodiment in which some or all of the components of the server 20 are incorporated into the communication device 10.

[0094] For example, in the above embodiments, it is also possible to implement an embodiment in which the structure and operation of the server 20 are distributed among multiple computers capable of communicating with each other. Additionally, for example, it is also possible to implement an embodiment in which some or all of the components of the server 20 are incorporated into the communication device 10.

[0095] In the above embodiment, the storage unit 25 may also store a database containing vehicle-related information. Alternatively, the first conversational model may be a RAG (Retrieval Augmented Generation) model that uses the database containing vehicle-related information when generating an answer to a user's question. When determining whether the prediction time is longer than a predetermined baseline, the control unit 24 may also determine whether the first conversational model needs to refer to the database containing vehicle-related information stored in the storage unit 25 to generate an answer. Specifically, if the server 20 determines that the first conversational model does not need to refer to the database, the control unit 24 determines that the prediction time is short. On the other hand, if the server 20 determines that the first conversational model needs to refer to the database, the control unit 24 determines that the prediction time is long. Therefore, dialogue with the second conversational model can be conducted until the generation of an answer using the first conversational model via RAG is completed, shortening the user's perceived time and ensuring the normal operation of the recognition system 1.

[0096] In addition, in the above embodiment, the server 20 may also send the question content related to the question to the communication device 10. Furthermore, the communication device 10 may display the question content on the output unit 12 until it receives an answer to the question content from the server 20. Figure 6 This is an example diagram showing a screen including screen 50 (second screen) and a question list 60. Question list 60 includes user question 61 and user question 62. User question 61 is a question such as "What is forward-looking driving assistance?". User question 62 is a question such as "Play only the same music?". Here, Figure 4 User questions 41 and 43 are respectively related to Figure 6 The user's questions 61 and 62 correspond. Additionally, user question 61 has already been answered; when selecting user question 61, it will be displayed... Figure 4 The first dialogic model responds to question 42. In the second screen 50, the user engages in dialogue with the second dialogic model. This allows the user to immediately identify the question. Furthermore, it confirms that System 1 is functioning correctly. Additionally, in... Figure 6In the example, the user-inputted question "Play only the same music" is consistent with the question content "Play only the same music" in the answer generated by server 20 in question list 60, but they do not necessarily have to be identical. For example, server 20 may perform query transformation including natural language processing, and send only the information extracted from the question to communication device 10, and communication device 10 may only display the information extracted from the question on output unit 12. Alternatively, communication device 10 may also display the question generated by server 20 based on the information extracted from the question on output unit 12. In this case, the user can confirm whether an answer to the question based on the user's intent has been generated, so the question can be modified as needed. As a result, the time until the user receives an answer to the question based on their intent can be shortened.

[0097] Furthermore, in the above embodiments, Figure 7 This is an example of a screen displaying a dialogue with the first conversational model and a dialogue with the second conversational model in a way that is distinguishable to the user on screen 40. Figure 7 In this system, a cloud-shaped message box displays the dialogue with the second conversational model, a rectangular message box with annotations appearing from the left displays the dialogue with the first conversational model, and a rectangular message box with annotations appearing from the right displays the user's question. Distinguishing methods include changing the shape or color of the message box, changing the size, font, or color of the text, or changing the sentence ending, but are not limited to these; any method can be used. Thus, the user can immediately determine whether they are interacting with the first or second conversational model. Furthermore, the user can confirm that System 1 is functioning correctly.

[0098] Furthermore, in the above embodiment, the server 20 may also send a message indicating that the response from the first conversational model is about to be completed or has been completed during the execution of the dialogue between the user and the second conversational model. Additionally, the communication device 10 may also display the messages received from the server 20 on the output unit 12. For example, in Figure 7 In this context, a message box using the second conversational model is used to display a message such as "The answer appears to be complete. We look forward to your feedback!" The message display can also utilize the second conversational model's message box, employing any method such as fixed text display. Furthermore, methods indicating that the answer via the first conversational model is nearing or has been completed can include message display, sound output, or vibration. Therefore, for example, even if a user is unable to concentrate on answering while driving, they can move to a safe location after the answer is generated to confirm it. Thus, the user can meaningfully pass the waiting time until the answer is generated.

[0099] Furthermore, in the above embodiment, the server 20 can also share the context of the user's dialogue with the first conversational model with the second conversational model. Regarding the method of sharing the context, for example, the control unit 24 can input information extracted from the user's question into the second conversational model, but it is not limited to this; any method can be used. By sharing the context of the first conversational model, the second conversational model can provide topics of interest to the user. Additionally, it can provide information beneficial to the user. Therefore, the user's perceived time can be shortened. Furthermore, since dialogue can take place on topics of interest to the user, user satisfaction with System 1 can be improved.

[0100] Furthermore, it is also possible to implement an embodiment in which a general-purpose computer functions as the server 20 described in the above embodiments. Specifically, a program describing the processing content for implementing the functions of the server 20 described in the above embodiments is stored in the memory of a general-purpose computer, and the program is read and executed by a processor. Therefore, this disclosure can also be implemented as a processor-executable program or a non-transitory computer-readable medium storing the program.

[0101] The following examples illustrate a portion of the embodiments of this disclosure. However, it is to be noted that the embodiments of this disclosure are not limited to these.

[0102] [Postscript 1]

[0103] A system comprising a communication device and a server, wherein,

[0104] The server stores the first dialogic model and the second dialogic model.

[0105] The communication device:

[0106] The first screen displays the dialogue between the user and the first conversational model.

[0107] The question text entered by the user on the first screen is sent to the server.

[0108] The server:

[0109] Determine whether the prediction time required for the first conversational model to generate an answer to the question related to the question text is longer than a predetermined benchmark.

[0110] If it is determined that the predicted time is longer than the predetermined benchmark, the dialogue between the user and the second conversational model is executed until the generation of the response through the first conversational model is completed.

[0111] [Postscript 2]

[0112] According to the system described in Appendix 1, in which,

[0113] The server:

[0114] The storage includes historical information comprising one or more previously entered questions and one or more answers to those questions.

[0115] If the question content related to the question input by the user is included in the historical information, it is determined that the prediction time is short, and the first conversational model is used to extract the answer content corresponding to the question content from the historical information. Based on the extracted answer content, an answer to the question content is generated.

[0116] If the question content related to the question input by the user is not included in the historical information, it is determined that the prediction time is long, and the first conversational model is used to generate an answer to the question content.

[0117] [Postscript 3]

[0118] According to the system described in Appendix 1 or 2, in which,

[0119] The server:

[0120] A database that stores information related to vehicles.

[0121] Determine whether the first conversational model needs to refer to the database in order to generate an answer related to the question text input by the user.

[0122] If it is determined that the first conversational model does not need to refer to the database, then the prediction time is short, and the first conversational model is used to generate an answer corresponding to the question.

[0123] If it is determined that the first conversational model needs to refer to the database, and the prediction time is long, then the first conversational model is used, referring to the database, to generate an answer corresponding to the question content.

[0124] [Postscript 4]

[0125] According to the system described in any of Appendix 1 to 3, wherein,

[0126] When the communication device is executing a dialogue between the user and the second conversational model, it overlays a second screen showing the dialogue with the second conversational model on the first screen.

[0127] [Postscript 5]

[0128] According to the system described in any of Appendix 1 to 4, wherein,

[0129] The server sends the question content related to the question text entered by the user to the communication device.

[0130] The communication device displays the question until an answer to the question is obtained from the server.

[0131] [Postscript 6]

[0132] According to the system described in any of Appendix 1 to 3 or 5, wherein,

[0133] When the communication device is executing a dialogue between the user and the second conversational model, it displays the dialogue with the first conversational model and the dialogue with the second conversational model in the first screen in a manner that the user can distinguish.

[0134] [Postscript 7]

[0135] According to the system described in any of Appendix 1 to 6, wherein,

[0136] During the execution of the dialogue between the user and the second conversational model, when the generation of the response from the first conversational model is completed, the server sends a message indicating that the response from the first conversational model is about to be completed or has been completed to the communication device.

[0137] The communication device displays the received message.

[0138] [Postscript 8]

[0139] A method, performed by a system equipped with communication devices and a server, wherein the method includes:

[0140] The server stores the first dialog model and the second dialog model;

[0141] The communication device:

[0142] The first screen displays the dialogue between the user and the first conversational model;

[0143] The question text entered by the user on the first screen is sent to the server;

[0144] The server:

[0145] Determine whether the prediction time required for the first conversational model to generate an answer to the question content related to the question text is longer than a predetermined benchmark; and

[0146] If it is determined that the predicted time is longer than the predetermined benchmark, the dialogue between the user and the second conversational model is executed until the generation of the response through the first conversational model is completed.

[0147] [Postscript 9]

[0148] According to the method described in Appendix 8, it also includes:

[0149] The server:

[0150] The storage includes historical information comprising one or more previously entered questions and one or more answers to those questions.

[0151] If the question content related to the question input by the user is included in the historical information, it is determined that the prediction time is short, and the first conversational model is used to extract the answer content corresponding to the question content from the historical information. Based on the extracted answer content, an answer to the question content is generated.

[0152] If the question content related to the question input by the user is not included in the historical information, it is determined that the prediction time is long, and the first conversational model is used to generate an answer to the question content.

[0153] [Postscript 10]

[0154] According to the method described in Appendix 8 or 9, it also includes:

[0155] The server:

[0156] A database containing vehicle-related information; and

[0157] Determine whether the first conversational model needs to refer to the database in order to generate an answer related to the question text input by the user.

[0158] If it is determined that the first conversational model does not need to refer to the database, then the prediction time is short, and the first conversational model is used to generate an answer corresponding to the question.

[0159] If it is determined that the first conversational model needs to refer to the database, and the prediction time is long, then the first conversational model is used, referring to the database, to generate an answer corresponding to the question content.

[0160] [Postscript 11]

[0161] The method described in any of the appendices 8-10, which also includes:

[0162] When the communication device is executing a dialogue between the user and the second conversational model, it overlays a second screen showing the dialogue with the second conversational model on the first screen.

[0163] [Postscript 12]

[0164] The method described in any of the appendices 8-11, which also includes:

[0165] The server sends the question content related to the question text input by the user to the communication device; and

[0166] The communication device displays the question until an answer to the question is obtained from the server.

[0167] [Postscript 13]

[0168] According to the method described in any of Appendix 8-10 or 12, wherein,

[0169] When the communication device is executing a dialogue between the user and the second conversational model, it displays the dialogue with the first conversational model and the dialogue with the second conversational model in the first screen in a manner that the user can distinguish.

[0170] [Postscript 14]

[0171] The method described in any of the appendices 8-13, which also includes:

[0172] During the execution of the dialogue between the user and the second conversational model, when the generation of the response from the first conversational model is completed, the server sends a message indicating that the response from the first conversational model is about to be completed or has been completed to the communication device; and

[0173] The communication device displays the received message.

[0174] [Postscript 15]

[0175] A program that causes a system equipped with communication devices and a server to perform actions including the following steps:

[0176] The server stores the first dialog model and the second dialog model;

[0177] The communication device:

[0178] The first screen displays the dialogue between the user and the first conversational model;

[0179] The question text entered by the user on the first screen is sent to the server;

[0180] The server:

[0181] Determine whether the prediction time required for the first conversational model to generate an answer to the question content related to the question text is longer than a predetermined benchmark; and

[0182] If it is determined that the predicted time is longer than the predetermined benchmark, the dialogue between the user and the second conversational model is executed until the generation of the response through the first conversational model is completed.

[0183] [Postscript 16]

[0184] According to the procedure described in Appendix 15, in which,

[0185] The action also includes:

[0186] The server:

[0187] The storage includes historical information comprising one or more previously entered questions and one or more answers to those questions.

[0188] If the question content related to the question input by the user is included in the historical information, it is determined that the prediction time is short, and the first conversational model is used to extract the answer content corresponding to the question content from the historical information. Based on the extracted answer content, an answer to the question content is generated.

[0189] If the question content related to the question input by the user is not included in the historical information, it is determined that the prediction time is long, and the first conversational model is used to generate an answer to the question content.

[0190] [Postscript 17]

[0191] According to the procedure described in Appendix 15 or 16, in which,

[0192] The action also includes:

[0193] The server:

[0194] A database containing vehicle-related information; and

[0195] Determine whether the first conversational model needs to refer to the database in order to generate an answer related to the question text input by the user.

[0196] If it is determined that the first conversational model does not need to refer to the database, then the prediction time is short, and the first conversational model is used to generate an answer corresponding to the question.

[0197] If it is determined that the first conversational model needs to refer to the database, and the prediction time is long, then the first conversational model is used, referring to the database, to generate an answer corresponding to the question content.

[0198] [Postscript 18]

[0199] According to the procedure described in any of the appendices 15-17, where,

[0200] The action also includes:

[0201] When the communication device is executing a dialogue between the user and the second conversational model, it overlays a second screen showing the dialogue with the second conversational model on the first screen.

[0202] [Postscript 19]

[0203] According to the procedure described in any of the appendices 15-17, where,

[0204] The action also includes:

[0205] The server sends the question content related to the question text input by the user to the communication device; and

[0206] The communication device displays the question until an answer to the question is obtained from the server.

[0207] When the user is performing a dialogue with the second conversational model, the dialogue with the first conversational model and the dialogue with the second conversational model are displayed in the first screen in a way that the user can distinguish.

[0208] [Postscript 20]

[0209] According to the procedure described in any of the appendices 15-19, where,

[0210] The action also includes:

[0211] During the execution of the dialogue between the user and the second conversational model, when the generation of the response from the first conversational model is completed, the server sends a message indicating that the response from the first conversational model is about to be completed or has been completed to the communication device; and

[0212] The communication device displays the received message.

[0213] (Symbol Explanation)

[0214] 1: System; 10: Communication device; 11: Communication unit; 12: Output unit; 13: Input unit; 14: Control unit; 15: Storage unit; 20: Server; 21: Communication unit; 22: Output unit; 23: Input unit; 24: Control unit; 25: Storage unit; 30: Network; 40: First screen; 41: User's question; 42: Answer of the first dialogic model; 43: User's question; 44: Animated icon; 50: Second screen; 60: Question list; 61: User's question; 62: User's question.

Claims

1. A system comprising a communication device and a server, wherein, The server stores the first dialogic model and the second dialogic model. The communication device: The first screen displays the dialogue between the user and the first conversational model. The question text entered by the user on the first screen is sent to the server. The server: Determine whether the prediction time required for the first conversational model to generate an answer to the question related to the question text is longer than a predetermined benchmark. If it is determined that the predicted time is longer than the predetermined benchmark, the dialogue between the user and the second conversational model is executed until the generation of the response through the first conversational model is completed.

2. The system according to claim 1, wherein, The server: The storage includes historical information comprising one or more previously entered questions and one or more answers to those questions. If the question content related to the question input by the user is included in the historical information, it is determined that the prediction time is short, and the first conversational model is used to extract the answer content corresponding to the question content from the historical information. Based on the extracted answer content, an answer to the question content is generated. If the question content related to the question input by the user is not included in the historical information, it is determined that the prediction time is long, and the first conversational model is used to generate an answer to the question content.

3. The system according to claim 1, wherein, The server: A database that stores information related to vehicles. Determine whether the first conversational model needs to refer to the database in order to generate an answer related to the question text input by the user. If it is determined that the first conversational model does not need to refer to the database, then the prediction time is short, and the first conversational model is used to generate an answer corresponding to the question. If it is determined that the first conversational model needs to refer to the database, and the prediction time is long, then the first conversational model is used, referring to the database, to generate an answer corresponding to the question content.

4. The system according to claim 1, wherein, When the communication device is executing a dialogue between the user and the second conversational model, it overlays a second screen showing the dialogue with the second conversational model on the first screen.

5. The system according to claim 1, wherein, The server sends the question content related to the question text entered by the user to the communication device. The communication device displays the question until an answer to the question is obtained from the server.

6. The system according to claim 1, wherein, When the communication device is executing a dialogue between the user and the second conversational model, it displays the dialogue with the first conversational model and the dialogue with the second conversational model in the first screen in a manner that the user can distinguish.

7. The system according to any one of claims 1 to 6, wherein, During the execution of the dialogue between the user and the second conversational model, when the generation of the response from the first conversational model is completed, the server sends a message indicating that the response from the first conversational model is about to be completed or has been completed to the communication device. The communication device displays the received message.

8. A method performed by a system equipped with communication devices and a server, wherein, The method includes: The server stores the first dialog model and the second dialog model; The communication device: The first screen displays the dialogue between the user and the first conversational model; The question text entered by the user on the first screen is sent to the server; The server: Determine whether the prediction time required for the first conversational model to generate an answer to the question content related to the question text is longer than a predetermined benchmark; and If it is determined that the predicted time is longer than the predetermined benchmark, the dialogue between the user and the second conversational model is executed until the generation of the response through the first conversational model is completed.

9. The method according to claim 8, wherein, Also includes: The server: The storage includes historical information comprising one or more previously entered questions and one or more answers to those questions. If the question content related to the question input by the user is included in the historical information, it is determined that the prediction time is short, and the first conversational model is used to extract the answer content corresponding to the question content from the historical information. Based on the extracted answer content, an answer to the question content is generated. If the question content related to the question input by the user is not included in the historical information, it is determined that the prediction time is long, and the first conversational model is used to generate an answer to the question content.

10. The method according to claim 8, wherein, Also includes: The server: A database containing vehicle-related information; and Determine whether the first conversational model needs to refer to the database in order to generate an answer related to the question text input by the user. If it is determined that the first conversational model does not need to refer to the database, then the prediction time is short, and the first conversational model is used to generate an answer corresponding to the question. If it is determined that the first conversational model needs to refer to the database, and the prediction time is long, then the first conversational model is used, referring to the database, to generate an answer corresponding to the question content.

11. The method according to claim 8, wherein, Also includes: When the communication device is executing a dialogue between the user and the second conversational model, it overlays a second screen showing the dialogue with the second conversational model on the first screen.

12. The method according to claim 8, wherein, Also includes: The server sends the question content related to the question text entered by the user to the communication device; as well as The communication device displays the question until an answer to the question is obtained from the server.

13. The method according to claim 8, wherein, When the communication device is executing a dialogue between the user and the second conversational model, it displays the dialogue with the first conversational model and the dialogue with the second conversational model in the first screen in a manner that the user can distinguish.

14. The method according to any one of claims 8 to 13, wherein, Also includes: During the execution of the dialogue between the user and the second conversational model, when the generation of the response from the first conversational model is completed, the server sends a message indicating that the response from the first conversational model is about to be completed or has been completed to the communication device; and The communication device displays the received message.

15. A non-transitory computer-readable medium storing a program that causes a system equipped with communication devices and a server to perform actions including the following steps: The server stores the first dialog model and the second dialog model; The communication device: The first screen displays the dialogue between the user and the first conversational model; The question text entered by the user on the first screen is sent to the server; The server: Determine whether the prediction time required for the first conversational model to generate an answer to the question content related to the question text is longer than a predetermined benchmark; and If it is determined that the predicted time is longer than the predetermined benchmark, the dialogue between the user and the second conversational model is executed until the generation of the response through the first conversational model is completed.

16. The non-transitory computer-readable medium according to claim 15, wherein, The action also includes: The server: The storage includes historical information comprising one or more previously entered questions and one or more answers to those questions. If the question content related to the question input by the user is included in the historical information, it is determined that the prediction time is short, and the first conversational model is used to extract the answer content corresponding to the question content from the historical information. Based on the extracted answer content, an answer to the question content is generated. If the question content related to the question input by the user is not included in the historical information, it is determined that the prediction time is long, and the first conversational model is used to generate an answer to the question content.

17. The non-transitory computer-readable medium according to claim 15, wherein, The action also includes: The server: A database containing vehicle-related information; and Determine whether the first conversational model needs to refer to the database in order to generate an answer related to the question text input by the user. If it is determined that the first conversational model does not need to refer to the database, then the prediction time is short, and the first conversational model is used to generate an answer corresponding to the question. If it is determined that the first conversational model needs to refer to the database, and the prediction time is long, then the first conversational model is used, referring to the database, to generate an answer corresponding to the question content.

18. The non-transitory computer-readable medium according to claim 15, wherein, The action also includes: When the communication device is executing a dialogue between the user and the second conversational model, it overlays a second screen showing the dialogue with the second conversational model on the first screen.

19. The non-transitory computer-readable medium according to claim 15, wherein, The action also includes: The server sends the question content related to the question text input by the user to the communication device; and The communication device displays the question until an answer to the question is obtained from the server. When the user is performing a dialogue with the second conversational model, the dialogue with the first conversational model and the dialogue with the second conversational model are displayed in the first screen in a way that the user can distinguish.

20. The non-transitory computer-readable medium according to any one of claims 15-19, wherein, The action also includes: During the execution of the dialogue between the user and the second conversational model, when the generation of the response from the first conversational model is completed, the server sends a message indicating that the response from the first conversational model is about to be completed or has been completed to the communication device; and The communication device displays the received message.

Images