JP2000207214A - Interaction device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide the interaction device which not only decides the feelings of a user and generates an answer to them, but also properly answers according to individual information of each user, the current state and the contents of interaction. SOLUTION: An interacting process means 17 determines the meaning structure and answer pattern of an answer output, on the basis of an input meaning structure and a feeling decision and extracts information needed to answer. When the input meaning structure clearly requests information, the information is obtained by referring to a knowledge base 12, an information server 10 and an individual information storage means 9. The obtained information is applied to a feeling inference rule, and the feelings of the user is estimated with the information according to the individual taste in an individual profile 9a. Based on the basis of the predicted user's feelings and the current feelings of the user, a proper answer pattern is selected to generate an output meaning structure.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an interactive device, and more particularly to an interactive device in which a user and a computer interact with each other through voice or facial expression.

[0002]

2. Description of the Related Art Conventionally, there has been known an interactive device for controlling a response on a computer side according to a user's emotion (Japanese Patent Laid-Open No. 8-339446). This conventional dialogue device includes a plurality of input means corresponding to a user's action or action, a feature extraction means for extracting features of an input signal input from the input means, and a plurality of signal features extracted by the feature extraction means. The apparatus includes an emotion determination unit that determines the emotion of the user, and a response generation unit that generates a response content of the computer based on the emotion determined by the emotion determination unit.

According to this conventional interactive device, it is possible to control the computer to respond according to the emotion of the user, so that a more familiar interactive device can be provided.

[0004]

However, in the above-described conventional dialogue apparatus, the computer only generates a response based on the judgment of the user's emotion based on the previously stored feature data, and the response feature reflecting the content of the dialogue. Does not. In addition, the cause of the emotion and the manner of appropriate response are not fixed by the user and the situation at the time,
In the above-described conventional interactive device, the response is determined only by the emotion determined based on the characteristics of the superficial input signal, and an appropriate response has not been achieved in response to these situations.

[0005] The present invention has been made in view of the above points,
It is an object of the present invention to provide a dialogue device that can not only determine a user's emotion and generate a response based on the emotion, but also perform an appropriate response according to personal information of each user, a situation at that time, and the content of the dialogue.

[0006]

SUMMARY OF THE INVENTION In order to achieve the above object, the present invention provides a dialogue device for interacting with a user through a voice input by the user or a facial expression of the user. An input unit for inputting at least linguistic information; a feature extraction unit for extracting a feature of the information input from the input unit; a feature extracted by the feature extraction unit; An emotion determining means for determining an emotion;
Input sentence analysis means for analyzing linguistic information input from the input means and extracting a semantic structure, information acquisition means for searching and acquiring information related to a topic, and personal information storing personal information for each user Storage means, emotion inference rule storage means for storing inference rules for associating objective matters with emotions, response pattern storage means for holding an index of a plurality of response patterns, and input analyzed by the input sentence analysis means The semantic structure of the linguistic information, the user's emotion determined by the emotion determination means, the related information obtained from the information obtaining means, the personal information obtained from the personal information storage means, and the emotion inference obtained from the emotion inference rule storage means The dialog processing means for selecting an appropriate response pattern from the response pattern storage means with reference to the rules and creating the semantic structure of the response sentence, Output sentence generation means for generating an output sentence from the generated semantic structure, response feature storage means for storing a response feature corresponding to the response pattern, output sentence generated by the output sentence generation means, and interactive processing means A configuration is provided that includes a response generation unit that generates a response content based on the response feature obtained from the response characteristic storage unit according to the selected response pattern, and an output unit that outputs the response content generated by the response generation unit. Things.

According to the present invention, the emotion of the user is determined based on the input information from the input means, the personal information of each user is accumulated, and the emotion inference rule is determined from the accumulated information and the information from the information acquisition means. By using, to estimate the cause of the emotion according to each user's preference, the situation at that time, the topic, predict the user's emotion to the response, create and output an appropriate response, it is easy to get familiar , An interactive device adapted to the user can be provided.

[0008]

Embodiments of the present invention will now be described with reference to the drawings. FIG. 1 is a block diagram showing an embodiment of a dialogue apparatus according to the present invention. As shown in FIG. 1, this embodiment includes an input unit 1 corresponding to a plurality of inputs from a user such as a voice and a face image, a feature extraction unit 2 for extracting a feature from an input signal of each input unit 1, Emotion feature storage means 3 for storing features corresponding to the emotion of the input signal
And emotion determination means 4 for determining the user's emotion from the characteristics of the signal obtained from the user's input based on the data stored in the emotion characteristic storage means 3, and output data for creating a response corresponding to the emotion. The response feature storing means 5, the response generating means 6 for generating the response content, and the output means 7 for outputting the generated data have the same configuration as the conventional interactive device.

In this embodiment, in addition to the configuration of the conventional apparatus, an input sentence analyzing means 8 for analyzing word information inputted from the input means 1 and extracting a meaning structure, and a personal information storage for storing personal information of a user. Means 9, an information server 10 for searching and acquiring information related to a topic, and an information searching and
Information acquisition means 13 comprising a communication means 11 and a knowledge base 12, emotion inference rule storage means 14 for storing inference rules for associating objective matters with user's emotions, and response pattern storage means for holding an index of a response pattern of a computer 15, a dialog history storage unit 16 for storing a dialog history, an input semantic structure, a user's emotion, related information acquired from the information acquisition unit 13, personal information, and an emotion inference rule to determine an appropriate response pattern. In this embodiment, a dialog processing unit 17 for selecting and creating a semantic structure of a response sentence and an output sentence generating unit 18 for generating an output word from the semantic structure are added.

The personal information storing means 9 stores a personal profile 9a. Schedule information 9b can be stored as shown in another embodiment. Examples of the use of the dialogue device described in the present embodiment include a device serving as a secretary, such as schedule management, information search and collection, and various information application devices. However, the dialogue device can be used for other purposes.

The input means 1 can include a plurality of input means corresponding to a plurality of inputs from a user such as voice, face image, sign language, and the like. The feature extracting unit 2 extracts a feature from an input signal of each input unit 1. These features include features related to the facial expression of the face image, voice pitch, and the like.
The emotion feature storage unit 3 stores a feature corresponding to the emotion of each input signal as data for determining the emotion. The emotion determination unit 4 determines the user's emotion from the characteristics of the input signal obtained from the user's input based on the data in the emotion characteristic storage unit 3.

The input sentence analyzing means 8 includes:
It analyzes information input as words, such as voice utterance, character input using a keyboard or pen, and sign language input, and creates an input meaning structure. The part omitted during the conversation is created by referring to the conversation history storage means 16 and supplementing the context.

The personal information storage means 9 stores personal information obtained in the course of the dialog for each user. The user profile is stored in the personal profile 9a. Full name·
In addition to address, age, gender, family composition, affiliation, status, etc., information on hobbies and preferences is accumulated through dialogue.

The emotion inference rule storage means 14 stores inference rules for causal relations regarding emotions. The inference rule is, for example, as shown in FIG. 2, the emotion "joy" inferred on the condition that a favorite thing or person gains profit (success).
(Pattern 1), emotion "disappointment" inferred on the condition that a favorite thing or person suffers a loss (failure), and inference on the condition that a disliked object or person gains (success) There is an emotion “anger” (pattern 3), and an emotion “satisfaction” (pattern 4) that is inferred on the condition that a disliked object or person suffers a loss (fails). There are individual differences in the amplitude of emotion and the degree of preference, but these are the individual profiles 9
The information in a is used. The degree to which a certain event satisfies the condition is evaluated with reference to information in the knowledge base 12.

The knowledge base 12 is a database relating to general knowledge and common sense. The information server 10 is a server that provides information in a distributed environment. It is conceivable to use a plurality of information servers 10 depending on the purpose. Examples of the information server 10 include a database of a specific field, multimedia information such as music and video, and the latest information supply such as news and weather forecast. The information search / communication means 11
The information server 1 receives a search request from the dialog processing means 17.
Select / connect 0 to search / acquire information.

The response pattern storage means 15 stores the response pattern of the computer. The response pattern can select a response content, a wording, and an index of various response characteristics. The conversation history storage means 16 is a means for accumulating the history of the conversation between the user and the computer. Even if the abbreviation is used in the user's spoken language, the conversation history storage means 16 refers to the conversation history to supplement the meaning and to make appropriate input. Required to create a semantic structure. Further, the dialog history storage unit 16 is referred to by the dialog processing unit 17 to create the next response based on the flow of the dialog.

The dialog processing means 17 determines the semantic structure and response pattern of the response output based on the input semantic structure and emotion determination, and extracts information necessary for a response. If the input semantic structure explicitly requests information, the information is acquired by referring to the knowledge base 12, the information server 10, and the personal information storage unit 9. The acquired information is applied to the emotion inference rule, and the user's emotion based on this information is estimated together with the personal preference in the personal profile 9a. An appropriate response pattern is selected based on the expected user's emotion and the current user's emotion, and an output semantic structure is created.

If the input semantic structure is not an explicit information request, and if the input semantic structure indicates the user's personal information,
The dialog processing means 17 stores the personal information in the personal information storage means 9. Using the personal profile 9a and the emotion inference rules based on the input semantic structure and the emotion determination, related information is acquired from the knowledge base 12 or the information server 10 as necessary to estimate the cause of the user's emotion. Dialogue processing means 17
Estimates the user's emotion based on the acquired information from the personal preference and emotion inference rules in the personal profile 9a, selects information to output and a response pattern, and creates an output semantic structure.

The output sentence generating means 18 includes the interactive processing means 17
A response sentence is generated based on the output semantic structure obtained from. The response sentence is expressed by a method desired by the user among text, voice, sign language, and the like. Response characteristic storage means 5 stores response characteristic data corresponding to the response pattern. The characteristics of the response include the expression and gesture of the character on the response screen, the tone of the response voice, and the like. The response generation means 6 extracts response characteristic data from the response characteristic storage means 5 based on the specified response pattern and generates a response output. The output means 7 outputs a response to a monitor screen, a speaker, or the like.

Next, a simple example of response generation according to the embodiment of the present invention will be described with reference to a case where professional baseball information is a topic. Actual emotion classification and inference rules are more fragmented and complicated, but simple examples are given for explanation. The emotion inference rule storage means 14 stores the inference rules shown in FIG. It is assumed that the user's personal profile 9a stores the user's preference (in this example, the name of a favorite professional baseball team and the name of a favorite player) as shown in FIG.

The knowledge base 12 stores general knowledge about baseball and semi-fixed knowledge. The information server 10 is updated with the latest information such as the match result and highlight scene of the day as needed. Information such as player registrations and game schedules is retrieved from the information server 10 as necessary, and the knowledge base 12
To update.

When the user A asks the interactive apparatus, "What is the result of today's match?"
Get more match results. If the result "Yokohama 5-1 giant" is obtained, "the user's favorite team has lost."
The dialog processing means 17 predicts that the user will be discouraged (estimates the emotion of the user), and outputs the information and the response pattern according to the predicted emotion. Is selected, an output semantic structure is created, response feature data is taken out from the response feature storage unit 5 based on the designated response pattern, a response output is generated by the response generation unit 6, and a response screen is generated by the output unit 7.
A response is output to a speaker or the like. As the response pattern, a pattern of answering unfortunately, "I'm sorry I lost."

On the other hand, when the user A says "Oh, lost again", it is assumed that the emotion of the user is determined to be discouraged from the voice / expression. In this case, when the match result is acquired from the information server 10 and the personal profile 9a is referred to, the cause of the user's disappointment is that the favorite team has lost, which matches the inference pattern 2. The interactive processing unit 17 of the computer searches for information that matches the inference pattern 1 that the user is likely to be pleased with as a response pattern, and causes the response generation unit 6 to create a response. In this case, "But
Matsui is the home run top. ", And a pattern that answers brightly is selected.

As described above, according to this embodiment, the personal information of each user is stored, and the stored information and the information from the knowledge base 12 and the information server 10 are used for each user using the emotion inference rule. It is possible to estimate the cause of the emotion according to the user's preference, the situation at that time, and the topic, predict the user's emotion to the computer response, and create an appropriate response.

Next, another embodiment of the present invention will be described. In this other embodiment, the personal information storage means 9
In addition to the personal profile 9a, schedule information 9b
Is stored. Schedule information 9
b stores the user's schedule. If the input meaning structure indicates the user's schedule in the dialogue with the user, the information is stored in the schedule information 9b. Using the personal profile 9a and the emotion inference rules based on the input semantic structure and the emotion determination, extracting relevant information from the knowledge base 12 or the information server 10 as necessary, and estimating the cause of the user's emotion, the schedule is used. The user's schedule is also referred to from the information 9b to estimate the cause of the emotion. In this case, the preferences of each user, the situation at that time,
An appropriate response can be created by estimating the cause of the emotion according to the schedule as well as the topic, estimating the user's emotion to the computer's response.

[0026]

As described above, according to the present invention,
The user's emotions are determined based on the input information from the input means, the personal information of each user is stored, and the stored information and the information from the information acquisition means are used to determine each user's emotion using the emotion inference rule. Estimating the cause of emotions according to tastes, current situations, and topics, predicting the user's emotions in response, and creating and outputting an appropriate response, not only the user's emotions but also their emotions The most suitable dialogue can be made for each user according to the user's preference, the situation at that time, and the topic, which is the cause of the problem.

[Brief description of the drawings]

FIG. 1 is a block diagram of an embodiment of the present invention.

FIG. 2 is a diagram illustrating an example of an inference rule.

FIG. 3 is a diagram illustrating an example of a personal profile.

[Explanation of symbols]

 Reference Signs List 1 input means 2 feature extraction means 3 emotion feature storage means 4 emotion determination means 5 response feature storage means 6 response generation means 7 output means 8 input sentence analysis means 9 personal information storage means 9a personal profile 9b schedule information 10 information server 11 information search Communication means 12 Knowledge base 13 Information acquisition means 14 Emotion inference rule storage means 15 Response pattern storage means 16 Dialog history storage means 17 Dialog processing means 18 Output sentence generation means

Claims (4)

[Claims] 1. An interactive device for interacting with a user through a voice input by the user or a facial expression of the user, etc., input means for inputting at least voice, face image, and language information issued by the user; and the input means A feature extraction unit that extracts a feature of information input from the device; an emotion determination unit that determines the user's emotion from the feature extracted by the feature extraction unit and emotion determination data stored in advance; An input sentence analyzing means for analyzing the linguistic information input from the input means and extracting a semantic structure; an information obtaining means for searching and obtaining information related to a topic; and an individual storing personal information for each user. Information storage means; emotion inference rule storage means for storing inference rules relating objective matters to emotions; and indexing of a plurality of response patterns. Response pattern storage means for storing the input language information analyzed by the input sentence analysis means, the user's emotion determined by the emotion determination means, and related information acquired from the information acquisition means. Referring to the personal information acquired from the personal information storage means and the emotion inference rules acquired from the emotion inference rule storage means, and selecting an appropriate response pattern from the response pattern storage means, and changing the semantic structure of the response sentence. Dialog processing means for creating; output sentence generating means for generating an output sentence from the semantic structure created by the dialog processing means; response feature storage means for storing response features corresponding to the response pattern; An output sentence generated by the sentence generation means and a response acquired from the response feature storage means according to the response pattern selected by the interaction processing means. Based on the characteristics, and the response generation means for generating a response content, interactive device, characterized in that an output means for outputting a response content generated by said response generating means. 2. The apparatus according to claim 1, further comprising: a dialog history storing unit configured to store a dialog history, wherein the input sentence analyzing unit supplements a meaning using a dialog history stored in the dialog history storing unit to create a semantic structure; 2. The interaction apparatus according to claim 1, wherein the interaction processing means selects the response pattern by referring to the interaction history stored in the interaction history storage means and creates a semantic structure of the response sentence. . 3. The interactive apparatus according to claim 1, wherein said personal information storage means holds at least one of a personal profile and schedule information for each user. 4. The information obtaining means receives an information server for providing desired latest information, a knowledge base storing data on general knowledge and common sense, and a search request from the interactive processing means. 2. The interactive apparatus according to claim 1, further comprising an information search / communication means for selecting / connecting the information server and acquiring desired information.