Interactive system, interactive method, two-way interactive system, two-way interactive method and recording medium

• US 7,127,398 B1
• Filed: 10/27/2000
• Issued: 10/24/2006
• Est. Priority Date: 10/29/1999
• Status: Expired due to Fees

• Current Assignee: ADIN Research Incorporated
• Original Assignee: ADIN Research Incorporated
• Inventors: Murata, Toshihiko, Fujimaki, Toshihide, Yamagishi, Kiichiro, Nakamura, Takeshi
• Primary Examiner(s): Dorvil, Richemond
• Assistant Examiner(s): Han, Qi
• Application Number:
  US09/697,146
• Time in Patent Office: 2,188 Days
• Field of Search: 704/270, 704/275, 704/231, 704/257
• US Class Current: 704/270
• CPC Class Codes: G10L 15/1815 Semantic context, e.g. disambiguation of the recognition hypotheses based on word meaning
• CPC Class Codes: G10L 15/1822 Parsing for meaning understanding
• CPC Class Codes: G10L 15/22 Procedures used during a speech recognition process, e.g. man-machine dialogue

Abstract

The present invention includes a microphone 1 inputting voice uttered by a user; a voice recognition processing section 3 recognizing the voice inputted by the microphone 1, and converting the recognized voice into an input symbol string; a conversation pattern processing section 5 inputting the input symbol string from the voice recognition processing section 3, and outputting an output symbol string corresponding to the input symbol string based on a conversation pattern described in advance; a voice synthesis processing section 7 converting the output symbol string from the conversation pattern processing section 5 into voice; a speaker 9 outputting and uttering the voice from the voice synthesis processing section 7; and a conversation learning section 41 grasping conversation characteristics of the user based on the input symbol string from the conversation pattern processing section 5, and changing the output symbol string in accordance with the grasped conversation characteristics.

Claims

1. An interactive system comprising:

a voice input section inputting voice uttered by a user;

a voice recognition processing section recognizing the voice inputted by the voice input section, and converting the recognized voice into an input symbol string;

a conversation pattern processing section inputting said input symbol string from the voice recognition processing section, and outputting an output symbol string corresponding to said input symbol string based on a conversation pattern described in advance;

a voice synthesis processing section converting said output symbol string from the conversation pattern processing section into voice;

a voice output section outputting and uttering the voice from the voice synthesis processing section; and

a conversation characteristic control section grasping conversation characteristics of said user and changing said output symbol string in accordance with the grasped conversation characteristics, the grasped conversation characteristics based on the continuity of conversation with respect to a specific topic,

wherein said conversation characteristic control section optimizes a conversation topic, a conversation pattern, uttered words, a conversation response interval, utterance frequency and timing during the utterance based on words and a conversation pattern used by the user with high frequency, a specific topic, times of conversation responses and utterance time and the number of words uttered by the user.

2. A computer readable recording medium for recording a program for allowing a computer to execute, comprising:

a voice recognition processing step of recognizing voice uttered by a user, and of converting the recognized voice into an input symbol string;

a conversation pattern processing step of inputting said input symbol string, and of outputting an output symbol string corresponding to said input symbol string based on a conversation pattern described in advance;

a voice synthesis processing step of converting said output symbol string into voice;

a voice output step of outputting and uttering the voice obtained in the voice synthesis processing step; and

a conversation characteristic control step of grasping conversation characteristics of said user and changing said output symbol string in accordance with the grasped conversation characteristics, the grasped conversation characteristics based on the continuity of conversation with respect to a specific topic,

wherein said conversation characteristic control section optimizes a conversation topic, a conversation pattern, uttered words, a conversation response interval, utterance frequency and timing during the utterance based on words and a conversation pattern used by the user with high frequency, a specific topic, times of conversation responses and utterance time and the number of words uttered by the user.

3. An interactive system comprising:

a voice input section inputting voice uttered by a user;

a voice recognition processing section recognizing the voice inputted by the voice input section, and converting the recognized voice into an input symbol string;

a conversation pattern processing section inputting said input symbol string from the voice recognition processing section, and outputting an output symbol string corresponding to said input symbol string based on a conversation pattern described in advance;

a voice synthesis processing section converting said output symbol string from the conversation pattern processing section into voice;

a voice output section outputting and uttering the voice from the voice synthesis processing section;

a storage section storing synonyms and relevant words synonymous with and relevant to words uttered by said user;

an association function section extracting the synonyms and relevant words synonymous with and relevant to words uttered by said user from said storage section based on a conversation history and said input symbol string from said conversation pattern processing section, and outputting said extracted synonyms and relevant words to said conversation pattern processing section,

wherein said extracted synonyms and relevant words are referred to by the conversation pattern processing section as variables when outputting the output symbol string; and

a conversation characteristic control section grasping conversation characteristics of said user and changing said output symbol string in accordance with the grasped conversation characteristics, the grasped conversation characteristics based on the continuity of conversation with respect to a specific topic,

wherein said conversation characteristic control section optimizes a conversation topic, a conversation pattern, uttered words, a conversation response interval, utterance frequency and timing during the utterance based on words and a conversation pattern used by the user with high frequency, a specific topic, times of conversation responses and utterance time and the number of words uttered by the user.

4. An interactive system comprising:

a voice input section inputting voice uttered by a user;

a first voice recognition processing section recognizing the voice inputted by the voice input section, and converting the recognized voice into a first input symbol string;

a conversation pattern processing section inputting said first input symbol string from the first voice recognition processing section, and outputting an output symbol string corresponding to said first input symbol string based on a conversation pattern described in advance;

a voice synthesis processing section converting said output symbol string from the conversation pattern processing section into voice;

a voice output section outputting and uttering the voice from the voice synthesis processing section;

a second recognition processing section recognizing broadcast voice from a broadcast station, and converting the recognized broadcast voice into second input symbol strings;

an event data storage section storing said second input symbol strings as event data in association with corresponding attribute identifiers; and

a conversation characteristic control section grasping conversation characteristics of said user and changing said output symbol string in accordance with the grasped conversation characteristics, the grasped conversation characteristics based on the continuity of conversation with respect to a specific topic,

wherein said conversation characteristic control section optimizes a conversation topic, a conversation pattern, uttered words, a conversation response interval, utterance frequency and timing during the utterance based on words and a conversation pattern used by the user with high frequency, a specific topic, times of conversation responses and utterance time and the number of words uttered by the user.

5. An interactive system according to claim 4, comprising an event data processing section extracting only a symbol string desired by the user among said second input symbol strings converted by said second voice recognition section, and storing the extracted symbol string in said event data storage section.

6. An interactive system according to claim 4, comprising:

a conversation characteristic control section grasping conversation characteristics of said user based on said input symbol string from said conversation pattern processing section, and changing said output symbol string in accordance with the grasped conversation characteristics.

7. An interactive system according to claim 4, comprising:

a storage section storing synonyms and relevant words synonymous with and relevant to words uttered by said user; and

an association function section extracting the synonyms and relevant words synonymous with and relevant to words uttered by said user from said storage section based on said input symbol string from said conversation pattern processing section, and outputting said extracted synonyms and relevant words to said conversation pattern processing section.

8. A two-way interactive system comprising a plurality of interactive systems each conducting an interactive processing for interaction with a user, the respective interactive systems communicating with one another, each of said interactive systems comprising:

a voice input section inputting voice uttered by a user;

a voice recognition processing section recognizing the voice inputted by the voice input section, and converting the recognized voice into an input symbol string;

a conversation pattern processing section inputting said input symbol string from the voice recognition processing section, and outputting an output symbol string corresponding to said input symbol string based on a conversation pattern described in advance;

a voice synthesis processing section converting said output symbol string from the conversation pattern processing section into voice;

a voice output section outputting and uttering the voice from the voice synthesis processing section;

an event data transmission and reception section transmitting said input symbol string to the other interactive system as event data, and receiving event data consisting of a symbol string from said other interactive system;

an event data storage section storing said received event data in association with corresponding attribute identifiers;

an input section inputting a privacy protection command of said user; and

a privacy protecting function section transmitting said event data acquired by utterance of said user to said other interactive system if said privacy protection command is not inputted from the input section, and not transmitting said event data acquired by the utterance of said user to said other interactive system if said privacy protection command is inputted from the input section.

9. A two-way interactive system according to claim 8, comprising:

a conversation characteristic control section grasping conversation characteristics of said user based on said input symbol string from said conversation pattern processing section, and changing said output symbol string in accordance with the grasped conversation characteristics.

10. A two-way interactive system according to claim 8, comprising:

a storage section storing synonyms and relevant words synonymous with and relevant to words uttered by said user; and

an association function section extracting the synonyms and relevant words synonymous with and relevant to words uttered by said user from said storage section based on said input symbol string from said conversation pattern processing section, and outputting said extracted synonyms and relevant words to said conversation pattern processing section.

11. A two-way interactive system according to claim 8, wherein a second recognition processing section recognizing broadcast voice from a broadcast station, and converting the recognized broadcast voice into second input symbol strings;

and

an event data storage section storing said second input symbol strings as event data.

12. An interactive method comprising:

a voice recognition processing step of recognizing voice uttered by a user, and of converting the recognized voice into an input symbol string;

a conversation pattern processing step of inputting said input symbol string, and of outputting an output symbol string corresponding to said input symbol string based on a conversation pattern described in advance in a conversation pattern processing section;

a voice synthesis processing step of converting said output symbol string into voice;

a voice output step of outputting and uttering the voice obtained in the voice synthesis processing step;

a storage step of storing synonyms and relevant words synonymous with and relevant to words uttered by said user;

an association function step of extracting the synonyms and relevant words synonymous with and relevant to words uttered by said user from said storage section based on a conversation history and said input symbol string, and of outputting said extracted synonyms and relevant words to said conversation pattern processing section,

wherein said extracted synonyms and relevant words are referred to by the conversation pattern processing section as variables when outputting the output symbol string; and

a conversation characteristic control step of grasping conversation characteristics of said user and changing said output symbol string in accordance with the grasped conversation characteristics, the grasped conversation characteristics based on the continuity of conversation with respect to a specific topic,

wherein said conversation characteristic control step optimizes a conversation topic, a conversation pattern, uttered words, a conversation response interval, utterance frequency and timing during the utterance based on words and a conversation pattern used by the user with high frequency, a specific topic, times of conversation responses and utterance time and the number of words uttered by the user.

13. An interactive method comprising:

a voice recognition processing step of recognizing voice uttered by a user, and of converting the recognized voice into an input symbol string;

a conversation pattern processing step of inputting said input symbol string, and of outputting an output symbol string corresponding to said input symbol string based on a conversation pattern described in advance;

a voice synthesis processing step of converting said output symbol string into voice;

a voice output step of outputting and uttering the voice obtained in the voice synthesis processing step; and

a conversation characteristic control step of grasping conversation characteristics of said user and changing said output symbol string in accordance with the grasped conversation characteristics, the grasped conversation characteristics based on the continuity of conversation with respect to a specific topic,

wherein said conversation characteristic control step optimizes a conversation topic, a conversation pattern, uttered words, a conversation response interval, utterance frequency and timing during the utterance based on words and a conversation pattern used by the user with high frequency, a specific topic, times of conversation responses and utterance time and the number of words uttered by the user.

14. An interactive method comprising:

a first voice recognition processing step of recognizing voice uttered by a user, and of converting the recognized voice into a first input symbol string;

a conversation pattern processing step of inputting said first input symbol string, and of outputting an output symbol string corresponding to said first input symbol string based on a conversation pattern described in advance;

a voice synthesis processing step of converting said output symbol string into voice;

a voice output section outputting and uttering the voice obtained in the voice synthesis processing step;

a second recognition processing step of recognizing broadcast voice from a broadcast station, and converting the recognized broadcast voice into second input symbol strings;

an event data storage step of storing said second input symbol strings as event data in association with corresponding attribute identifiers; and

a conversation characteristic control step of grasping conversation characteristics of said user and changing said output symbol string in accordance with the grasped conversation characteristics, the grasped conversation characteristics based on the continuity of conversation with respect to a specific topic,

wherein said conversation characteristic control step optimizes a conversation topic, a conversation pattern, uttered words, a conversation response interval, utterance frequency and timing during the utterance based on words and a conversation pattern used by the user with high frequency, a specific topic, times of conversation responses and utterance time and the number of words uttered by the user.

15. An interactive method according to claim 14, comprising:

an event data processing step of extracting only a symbol string desired by the user among said second input symbol strings converted in said second voice recognition step, and of storing the extracted symbol string in said event data storage section.

16. An interactive method according to claim 14, comprising:

a conversation characteristic control step of grasping conversation characteristics of said user based on said input symbol string, and of changing said output symbol string in accordance with the grasped conversation characteristics.

17. An interactive method according to claim 14, comprising:

a storage step of storing synonyms and relevant words synonymous with and relevant to words uttered by said user; and

an association function step of extracting the synonyms and relevant words synonymous with and relevant to words uttered by said user from said storage section based on said input symbol string, and of outputting said extracted synonyms and relevant words to said conversation pattern processing section.

18. A two-way interactive method of allowing a plurality of interactive systems each conducting an interactive processing for interaction with a user to communicate with one another, the method comprising:

a voice recognition processing step of recognizing the voice uttered by a user, and of converting the recognized voice into an input symbol string;

a conversation pattern processing step of inputting said input symbol string, and of outputting an output symbol string corresponding to said input symbol string based on a conversation pattern described in advance;

a voice synthesis processing step of converting said output symbol string into voice;

a voice output section outputting and uttering the voice obtained in the voice synthesis processing step;

an event data transmission and reception step of transmitting said input symbol string to the other interactive system as event data, and of receiving event data consisting of the symbol string from said other interactive system;

an event data storage step of storing said received event data in association with corresponding attribute identifiers;

an input step of inputting a privacy protection command of said user; and

a privacy protecting function step of transmitting said event data acquired by utterance of said user to said other interactive system if said privacy protection command is not inputted from the input section, and not transmitting said event data acquired by the utterance of said user to said other interactive system if said privacy protection command is inputted from the input section.

19. A two-way interactive method according to claim 18, comprising:

a conversation characteristic control step of grasping conversation characteristics of said user based on said input symbol string, and of changing said output symbol string in accordance with the grasped conversation characteristics.

20. A two-way interactive method according to claim 18, comprising:

a storage step of storing synonyms and relevant words synonymous with and relevant to words uttered by said user; and

an association function step of extracting the synonyms and relevant words synonymous with and relevant to words uttered by said user from said storage section based on said input symbol string, and of outputting said extracted synonyms and relevant words to said conversation pattern processing section.

21. A two-way interactive method according to claim 18, comprising:

a second recognition processing step of recognizing broadcast voice from a broadcast station, and of converting the recognized broadcast voice into second input symbol strings; and

an event data storage step of storing said second input symbol strings as event data.