Dialogue device, dialogue method, and dialogue program
The dialogue device facilitates interactions between chatbot models with defined personalities by acquiring persona information, selecting appropriate models, and generating dialogues, addressing the lack of interaction and evaluation in conventional systems and reducing data collection costs.
Patent Information
- Authority / Receiving Office
- JP · JP
- Patent Type
- Patents
- Current Assignee / Owner
- NIPPON TELEGRAPH & TELEPHONE CORP
- Filing Date
- 2022-06-20
- Publication Date
- 2026-05-15
AI Technical Summary
Conventional chatbot dialogue models lack the ability to interact with each other, as the value of dialogue logs is not recognized, and their characteristics are unclear, making it difficult to evaluate their results quantitatively and qualitatively.
A dialogue device that acquires persona information for each model, selects a dialogue model based on this information, generates an initial utterance, and uses the selected model to generate response utterances, enabling dialogues between models with defined personalities.
Enables dialogues between chatbot dialogue models with defined personalities, facilitating communication simulations and reducing the cost of collecting dialogue data by automatically switching personas, allowing evaluation of conversation results.
Smart Images
Figure 0007859496000001 
Figure 0007859496000002 
Figure 0007859496000003
Abstract
Description
Technical Field
[0001] The present invention relates to an interaction device, an interaction method, and an interaction program.
Background Art
[0002] Conventionally, a chatbot that allows a person and a computer to interact through text or voice using an interaction model is known and has been introduced in various services. In recent years, with the development of deep learning technology, an interaction model capable of performing an interaction without a sense of strangeness at the same level as a human has been constructed (see Non-Patent Document 1).
Prior Art Documents
Non-Patent Documents
[0003]
Non-Patent Document 1
Summary of the Invention
Problems to be Solved by the Invention
[0004] However, conventional technologies do not involve chatbot dialogue models interacting with each other. This is because the value of dialogue logs from conversations between chatbots has not been recognized. Furthermore, since the dialogue models themselves are black boxes, even if multiple dialogue models are prepared, it is unclear what kind of characteristics each dialogue model possesses, making it difficult to quantitatively and qualitatively evaluate the results of dialogues between dialogue models.
[0005] This invention has been made in view of the above, and aims to enable dialogue between chatbot dialogue models, each with defined individual characteristics. [Means for solving the problem]
[0006] To solve the above-mentioned problems and achieve the objective, the dialogue device according to the present invention is characterized by comprising: an acquisition unit that acquires persona information representing the individuality of each dialogue model to be used for dialogue; a selection unit that selects a dialogue model corresponding to the persona information; a generation unit that generates an initial utterance to initiate a dialogue; and a dialogue generation unit that uses the selected dialogue model to generate a response utterance to the generated initial utterance. [Effects of the Invention]
[0007] According to the present invention, it becomes possible for chatbot dialogue models, each with a defined personality, to engage in dialogue with one another. [Brief explanation of the drawing]
[0008] [Figure 1] Figure 1 is a schematic diagram illustrating the general configuration of the dialogue device of this embodiment. [Figure 2] Figure 2 is a diagram illustrating the processing of the dialogue device. [Figure 3] Figure 3 shows an example of the screen display of the dialogue device. [Figure 4] Figure 4 is a flowchart showing the dialogue processing procedure. [Figure 5] Figure 5 shows an example of a computer running an interactive program. [Modes for carrying out the invention]
[0009] Hereinafter, one embodiment of the present invention will be described in detail with reference to the drawings. However, the present invention is not limited to this embodiment. Furthermore, in the drawings, the same parts are denoted by the same reference numerals.
[0010] [Configuration of the dialogue device] Figure 1 is a schematic diagram illustrating the general configuration of the dialogue device of this embodiment. Figure 2 is a diagram illustrating the processing of the dialogue device. Figure 3 is a diagram showing an example of the screen display of the dialogue device. First, as illustrated in Figure 1, the dialogue device 10 of this embodiment is implemented using a general-purpose computer such as a personal computer, and includes an input unit 11, an output unit 12, a communication control unit 13, a storage unit 14, and a control unit 15.
[0011] The input unit 11 is implemented using an input device such as a keyboard or mouse, and in response to input operations by the operator, it inputs various instruction information, such as the start of processing, to the control unit 15. The output unit 12 is implemented using a display device such as a liquid crystal display, a printing device such as a printer, etc. For example, the output unit 12 displays the results of the interactive processing described later.
[0012] The communication control unit 13 is implemented using a NIC (Network Interface Card) or the like, and controls communication between the control unit 15 and external devices via telecommunication lines such as a LAN (Local Area Network) or the Internet. For example, the communication control unit 13 controls communication between the control unit 15 and a user terminal used by a user who performs the dialogue processing described later.
[0013] The memory unit 14 is implemented by semiconductor memory elements such as RAM (Random Access Memory) and flash memory, or by storage devices such as hard disks and optical discs. The memory unit 14 pre-stores processing programs for operating the dialogue device 10, and data used during the execution of the processing programs, or temporarily stores them each time processing is performed. The memory unit 14 may also be configured to communicate with the control unit 15 via the communication control unit 13. In this embodiment, the memory unit 14 stores a dialogue model 14a used in the dialogue processing described later.
[0014] The control unit 15 is implemented using a CPU (Central Processing Unit) or the like, and executes a processing program stored in memory. As a result, the control unit 15 functions as an acquisition unit 15a, a selection unit 15b, a generation unit 15c, a dialogue generation unit 15d, and a decision unit 15e, as illustrated in Figure 1, to perform dialogue processing. Note that each or part of these functional units may be implemented on different hardware. For example, the dialogue generation unit 15d may be implemented on different hardware than the other functional units. Furthermore, the control unit 15 may also include other functional units.
[0015] The acquisition unit 15a acquires persona information that represents the individuality of each dialogue model to be used for the conversation. Here, persona information is information that represents individuality, such as "gender - male" and "hobby - fishing," and is, for example, information described in key-value format text.
[0016] The acquisition unit 15a acquires user-defined persona information (PA, PB) for each of the two dialogue models that the user wants to interact with, for example, as shown in Figure 2. Specifically, the acquisition unit 15a acquires persona information (PA, PB) via the input unit 11 or via the communication control unit 13 from the user terminal operated by the user.
[0017] Further, FIG. 3 illustrates a screen presented to the user at the output unit 12 or the user terminal. In the example shown in FIG. 3, persona information is defined by the user inputting persona A definition and persona B definition.
[0018] Further, the acquisition unit 15a may acquire randomly determined persona information. For example, when the user presses the "random persona button" illustrated in FIGS. 2 and 3, the acquisition unit 15a acquires randomly determined persona information from the definable persona information.
[0019] In this case, for example, the determination unit 15e may determine, as persona information, the value corresponding to the randomly determined key using a list of candidate keys created in advance. Alternatively, the determination unit 15e may determine persona information by utilizing a model or algorithm that generates a character string for the free-form value with respect to the given random seed.
[0020] Further, the acquisition unit 15a acquires topic information for specifying a topic. Here, the topic information is information for specifying the content of the topic for which the dialogue model is to be made to dialogue, and is, for example, any character string information described in text. The acquisition unit 15a acquires the topic information T via the input unit 11 as shown in FIG. 2, for example, or from the user terminal operated by the user via the communication control unit 13. Also, in the example shown in FIG. 3, the topic information is specified by the user as a topic definition.
[0021] Returning to the description of FIG. 1, the selection unit 15b selects a dialogue model corresponding to the persona information. For example, the dialogue model 14a may be generated in advance corresponding to each of the persona information or combination of persona information that can be acquired by the acquisition unit 15a. In that case, the selection unit 15b selects the dialogue model corresponding to the persona information or combination of a plurality of persona information acquired by the acquisition unit 15a.
[0022] Alternatively, the dialogue model 14a may be pre-generated to correspond to each key-value of the persona information that the acquisition unit 15a can acquire. If the acquisition unit 15a acquires multiple pieces of persona information, the selection unit 15b may generate a dialogue model that corresponds to the combination of multiple pieces of persona information by selecting and integrating the dialogue models corresponding to each piece of persona information.
[0023] Alternatively, the persona information, such as a string acquired by the acquisition unit 15a, may be converted into a predetermined format by the selection unit 15b and input into the dialogue model 14a, thereby generating a dialogue model corresponding to the persona information.
[0024] The generation unit 15c generates an initial utterance to initiate a dialogue. Specifically, the generation unit 15c generates an initial utterance corresponding to topic information that specifies a topic. For example, if the topic information is the word XX, a template for the initial utterance, such as "What do you think about XX?", is prepared in advance. This allows the generation unit 15c to generate an initial utterance corresponding to the topic information. Alternatively, if the topic information is input in utterance format, the generation unit 15c may use the input topic information as the initial utterance.
[0025] The dialogue generation unit 15d generates a response utterance to the generated initial utterance using the selected dialogue model. For example, when generating a dialogue between dialogue models A and B, if the initial utterance of dialogue model A is generated, the dialogue generation unit 15d outputs a response utterance by inputting the initial utterance to dialogue model B, which is selected as the dialogue partner for dialogue model A.
[0026] The dialogue generation unit 15d uses the selected dialogue model to further generate response utterances to the generated response utterances. In other words, as illustrated in Figure 2, the dialogue generation unit 15d uses dialogue model A to generate response utterances to the input utterances, taking the response utterances generated for the initial utterance as input. The dialogue generation unit 15d also uses dialogue model B to generate response utterances to the input utterances, taking the response utterances generated for the response utterances as input. By repeating this process, the dialogue generation unit 15d generates dialogue between the dialogue models.
[0027] The dialogue generation unit 15d terminates the dialogue when predetermined termination conditions are met. For example, the dialogue generation unit 15d terminates the generation of the dialogue when it reaches a predetermined number of utterances set by hyperparameters. Alternatively, as illustrated in Figure 3, a "dialogue termination button" may be set, and the dialogue generation unit 15d may terminate the generation of the dialogue when the user presses the "dialogue termination button". Alternatively, the dialogue generation unit 15d may evaluate the continuity of the topic using a pre-built topic continuity determination model and terminate the generation of the dialogue when it determines that the topic has ended.
[0028] Furthermore, the dialogue generation unit 15d outputs the generated dialogue to the output unit 12 and presents it to the user. For example, as illustrated in Figure 3, it is output in chat format. Alternatively, the dialogue generation unit 15d may output the dialogue results in file format, including topic information and persona information. This makes it possible to simulate dialogues between dialogue models with defined personalities.
[0029] [Dialogue Processing] Next, with reference to Figure 4, the dialogue processing by the dialogue device 10 according to this embodiment will be described. Figure 4 is a flowchart of the dialogue processing procedure. The flowchart in Figure 4 starts, for example, when the user inputs an operation to instruct the start.
[0030] First, the acquisition unit 15a acquires persona information that represents the individuality of each dialogue model to be used for the conversation (step S1). For example, the acquisition unit 15a acquires persona information defined by the user for each of the two dialogue models that the user wants to have a conversation with. Alternatively, the acquisition unit 15a acquires randomly determined persona information.
[0031] Furthermore, the acquisition unit 15a acquires topic information to specify a topic.
[0032] Next, the selection unit 15b selects a dialogue model corresponding to the persona information (step S2). For example, the selection unit 15b selects a dialogue model corresponding to the acquired persona information from the dialogue models 14a that have been generated in advance to correspond to persona information or a combination of multiple persona information.
[0033] Furthermore, the generation unit 15c generates the initial utterance in response to the topic information (step S3).
[0034] Then, the dialogue generation unit 15d generates a response utterance to the input utterance using the selected dialogue model (step S4). First, the dialogue generation unit 15d takes the generated initial utterance as input and generates a response utterance to the initial utterance using the selected dialogue model.
[0035] Furthermore, the dialogue generation unit 15d takes the generated response utterance as input and uses the selected dialogue model to generate a response utterance for the input response utterance, repeating this process until a predetermined termination condition is met (step S5, No → step S4), thereby generating a dialogue. When the predetermined termination condition is met (step S5, Yes), the dialogue generation unit 15d outputs the generated dialogue to the output unit 12 (step S6), ending the series of dialogue processes.
[0036] [effect] As described above, in the dialogue device 10 of this embodiment, the acquisition unit 15a acquires persona information representing the individuality of each dialogue model to be used for dialogue. The selection unit 15b selects a dialogue model corresponding to the persona information. The generation unit 15c generates an initial utterance to start the dialogue. The dialogue generation unit 15d uses the selected dialogue model to generate a response utterance to the generated initial utterance.
[0037] Specifically, the acquisition unit 15a further acquires topic information that specifies the topic. In this case, the generation unit 15c generates the initial utterance corresponding to the topic information.
[0038] Furthermore, the dialogue generation unit 15d uses the selected dialogue model to generate further response utterances to the generated response utterances.
[0039] This enables dialogue between chatbot dialogue models with various defined personalities. Such dialogues between dialogue models can be used as a communication simulation. For example, by setting a target persona for a particular topic and having them engage in a dialogue, it becomes possible to check their reactions to that topic, which can then be used for marketing, user recommendations, and other applications.
[0040] Furthermore, by automatically switching personas, it becomes possible to collect numerous conversation patterns on the same topic. Therefore, it is possible to reduce the cost of collecting dialogue data without collecting it as actual dialogues. In this way, the dialogue device 10 makes it possible to evaluate the results of conversations between chatbot dialogue models.
[0041] Furthermore, the acquisition unit 15a may acquire randomly determined persona information. This makes it easy to collect dialogue data for an even larger number of patterns.
[0042] [program] It is also possible to create a program in a computer-executable language that describes the processing performed by the dialogue device 10 according to the above embodiment. In one embodiment, the dialogue device 10 can be implemented by installing the dialogue program that performs the above dialogue processing as packaged software or online software on a desired computer. For example, by having the above dialogue program run on an information processing device, the information processing device can be made to function as the dialogue device 10. The information processing device referred to here includes desktop or notebook personal computers. In addition, the category of information processing device also includes mobile communication terminals such as smartphones, mobile phones and PHS (Personal Handyphone System), and slate terminals such as PDA (Personal Digital Assistant). Furthermore, the functions of the dialogue device 10 may be implemented on a cloud server.
[0043] Figure 5 shows an example of a computer running an interactive program. Computer 1000 includes, for example, memory 1010, a CPU 1020, a hard disk drive interface 1030, a disk drive interface 1040, a serial port interface 1050, a video adapter 1060, and a network interface 1070. These components are connected by a bus 1080.
[0044] Memory 1010 includes ROM (Read Only Memory) 1011 and RAM 1012. ROM 1011 stores, for example, a boot program such as BIOS (Basic Input Output System). The hard disk drive interface 1030 is connected to the hard disk drive 1031. The disk drive interface 1040 is connected to the disk drive 1041. A removable storage medium, such as a magnetic disk or optical disk, is inserted into the disk drive 1041. A serial port interface 1050 is connected to, for example, a mouse 1051 and a keyboard 1052. A video adapter 1060 is connected to, for example, a display 1061.
[0045] Here, the hard disk drive 1031 stores, for example, the OS 1091, the application program 1092, the program module 1093, and the program data 1094. The information described in the above embodiment is stored, for example, in the hard disk drive 1031 or the memory 1010.
[0046] Furthermore, the dialogue program is stored in the hard disk drive 1031 as a program module 1093 containing instructions to be executed by the computer 1000, for example. Specifically, the program module 1093 containing instructions for each process executed by the dialogue device 10 described in the above embodiment is stored in the hard disk drive 1031.
[0047] Furthermore, the data used for information processing by the interactive program is stored as program data 1094, for example, in the hard disk drive 1031. The CPU 1020 then reads the program module 1093 and program data 1094 stored in the hard disk drive 1031 into the RAM 1012 as needed and executes the procedures described above.
[0048] Furthermore, the program module 1093 and program data 1094 related to the interactive program are not limited to being stored on the hard disk drive 1031; for example, they may be stored on a removable storage medium and read by the CPU 1020 via a disk drive 1041 or the like. Alternatively, the program module 1093 and program data 1094 related to the interactive program may be stored on another computer connected via a network such as a LAN or WAN (Wide Area Network) and read by the CPU 1020 via a network interface 1070.
[0049] Although embodiments applying the invention made by the present inventors have been described above, the present invention is not limited by the descriptions and drawings that constitute part of the disclosure of the present invention in this embodiment. That is, all other embodiments, examples, and operational techniques made by those skilled in the art based on this embodiment are included in the scope of the present invention. [Explanation of Symbols]
[0050] 10 Dialogue device 11 Input section 12 Output section 13 Communication Control Unit 14 Storage section 14a Dialogue Model 15 Control Unit 15a Acquisition part 15b Selection Section 15c Generator 15d Dialogue Generation Unit 15e Decision section
Claims
1. An acquisition unit that acquires multiple persona information representing the individuality of each dialogue model to be used in the conversation, A selection unit that selects and integrates dialogue models corresponding to each of the multiple pieces of persona information obtained, A generation unit that generates the initial utterance to initiate a dialogue, A dialogue generation unit that generates a response utterance to the generated initial utterance using the integrated dialogue model, A dialogue device characterized by having the following features.
2. The aforementioned acquisition unit further acquires information that specifies the topic, The generation unit generates the initial utterance in accordance with the information specifying the topic. The dialogue device according to feature 1.
3. The dialogue device according to claim 1, characterized in that the dialogue generation unit further generates response utterances to the generated response utterances using the dialogue model.
4. The dialogue device according to claim 1, characterized in that the acquisition unit acquires the randomly determined persona information.
5. A method of dialogue performed by a dialogue device, The acquisition process involves obtaining multiple persona information that represents the individuality of each dialogue model to be used in the conversation, A selection step of selecting and integrating dialogue models corresponding to each of the multiple persona information obtained, A generation process that generates the initial utterance to initiate a dialogue, A dialogue generation process that generates response utterances to the generated initial utterance using the integrated dialogue model, A dialogue method characterized by including the following.
6. The acquisition step involves obtaining multiple persona information that represents the individuality of each dialogue model to be used in the conversation, A selection step involves selecting and integrating a dialogue model corresponding to each of the multiple pieces of persona information obtained, A generation step that generates an initial utterance to initiate a dialogue, A dialogue generation step that generates a response utterance to the generated initial utterance using the integrated dialogue model, An interactive program that instructs a computer to execute a command.