Systems and methods for responding to natural language speech utterance

• US 20040044516A1
• Filed: 06/03/2003
• Published: 03/04/2004
• Est. Priority Date: 06/03/2002
• Status: Active Grant

• Current Assignee: Dialect, LLC
• Original Assignee: VoiceBox Technologies Corporation (Microsoft Corporation)
• Inventors: Kennewick, Robert A., Locke, David, Kennewick, Richard, Kennewick, Michael R. Jr., Freeman, Tom
• Granted Patent: US 7,398,209 B2
• Time in Patent Office: Days
• US Class Current: 704/5
• 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
• CPC Class Codes: G10L 2015/228 of application context
• CPC Class Codes: Y10S 707/99933 Query processing, i.e. searching

Abstract

Systems and methods for receiving natural language queries and/or commands and execute the queries and/or commands. The systems and methods overcomes the deficiencies of prior art speech query and response systems through the application of a complete speech-based information query, retrieval, presentation and command environment. This environment makes significant use of context, prior information, domain knowledge, and user specific profile data to achieve a natural environment for one or more users making queries or commands in multiple domains. Through this integrated approach, a complete speech-based natural language query and response environment can be created. The systems and methods creates, stores and uses extensive personal profile information for each user, thereby improving the reliability of determining the context and presenting the expected results for a particular question or command.

Claims

1. A system responsive to a user generated natural language speech utterance, comprising:

an event manager, the event manager coordinating interaction between components of the system;

a plurality of domain agents, the domain agents being autonomous executable that receives, processes and responds to at least one of a query or a command; and

a parser that determines a domain for the user generated natural language utterance based on the content and context of the user utterance

2. The system according to claim 1, further comprising a speech unit for receiving the user generated natural language utterance

3. The system according to claim 2, wherein the speech unit further comprises a speech code.

4. The system according to claim 1, wherein the utterance comprising at least one of a user query and a user command.

5. The system according to claim 1, further comprising a speech recognition engine that processes the user generated natural language utterance using data in a dictionary and phrase module.

6. The system according to claim 1, further comprising user profiles.

7. The system according to claim 1, further comprising a text to speech engine.

8. The system according to claim 1, further comprising a personality module that facilitates formatting of a response generated by the system.

9. The system according to claim 1, wherein the parser determines a context of the user utterance based on at least one of applying prior probabilities or fuzzy possibilities to keyword matching, user profile, and dialog history.

10. The system according to claim 1, wherein at least one of the domain agents creates and directs a query to at least one of a local information source and a network information source.

11. The system according to claim 1, wherein at least one of the domain agents creates and directs a command to the system or a remote or local device.

12. The system according to claim 10, wherein at least one of the domain agents evaluates multiple responses from multiple sources in response to the query.

13. The system according to claim 12, wherein at least one of the domain agents evaluates the multiple responses by scoring each response based on results already received, context of the user utterance, user utterance criteria, history of dialog between the user and the system, a user profile or domain specific information using probabilistic or fuzzy scoring techniques.

14. The system according to claim 1, wherein at least one of the domain agents formats generated responses using available format templates and based on user utterance context, user utterance criteria, or a user profile.

15. The system according to claim 14, wherein the responses are generated by performing variable substitutions and transformations to produce a response best understood and most natural to the user.

16. A method responsive to a user generated natural language speech utterance, comprising the steps of:

receiving the user generated natural language speech utterance,

determining a domain for the user utterance; and

selecting at least one domain agent for processing the user utterance based on the determined domain, the domain agent being an autonomous executable that receives, processes and responds to at least one of a query and a command.

17. The method according to claim 16, further comprising the step of processing the user utterance to determine its meaning including determining at least one query or command that is contained in the user utterance.

18. The method according to claim 17, wherein the step of determining the meaning of the user utterance includes determining a context of the utterance based on at least one of applying prior probabilities or fuzzy possibilities to keyword matching, user profile, and dialog history.

19. The method according to claim 18, wherein the domain is determined at least in part by the context.

20. The method according to claim 16, further comprising the step of generating at least one of a query and a command understandable by one domain agents.

21. The method according to claim 16, wherein the query or command is executed by at least one domain agent.

22. The method according to claim 21, wherein the at least one domain agent executes a query by submitting the query to a local information source or a network information source.

23. The method according to claim 21, wherein the at least one domain agent executes a command by submitting a command to the system or a remote or local device.

24. The method according to claim 22, wherein the at least one domain agent evaluates multiple responses from multiple sources.

25. The method according to claim 24, wherein the at least one domain agent evaluates the multiple responses by scoring each response based on results already received, context of the user utterance, user utterance criteria, history of dialog between the user and the system, a user profile or domain specific information using probabilistic or fuzzy scoring techniques.

26. The method according to claim 16, wherein the at least one domain agent formats generated responses using available format templates and based on user utterance context, user utterance criteria, or a user profile.

27. The method according to claim 26, wherein the responses are generated by performing variable substitutions and transformations to produce a response best understood and most natural to the user.

28. The method according to claim 16, wherein the user utterance is received from a speech unit.

29. The method according to claim 28, wherein the user utterance is coded by the speech unit and received from the speech unit via communication link.

30. A system for executing a user generated natural language speech utterance, the utterance having a query or a command, comprising:

receiving means that receives the user generated natural language speech utterance;

domain determining means that determines a domain for the user utterance; and

selecting means that selects at least one domain agent for processing the user utterance based on the determined domain that is autonomous executable that receives, processes and responds to at least one of a query and a command.

34. The system according to claim 30, further comprising a generating means for generating a query or a command understandable by the at least one domain agent.

35. The system according to claim 34, wherein the at least one domain agent having a means for executing a query by submitting the query to a local information source or a network information source.

36. The system according to claim 34, wherein the at least one domain agent having a means for executing a command by submitting a command the system or a remote or local device.

37. The system according to claim 34, wherein the at least one domain agent having a means for evaluating multiple responses from multiple sources.

38. The system according to claim 37, wherein the at least one domain agent having a means for evaluating the multiple responses by scoring each response based on results already received, context of the user utterance, user utterance criteria, history of dialog between the user and the system, a user profile or domain specific information using probabilistic or fuzzy scoring techniques.

39. The system according to claim 30, wherein the at least one domain agent having a means for formatting generated responses using available format templates and based on user utterance context, user utterance criteria, or a user profile.

40. The system according to claim 39, wherein the responses are generated by performing variable substitutions and transformations to produce a response best understood and most natural to the user.

31. The system according to claim 31, further comprising a means for processing the user utterance in order to determine its meaning including determining at least one of query and command that is contained in the user utterance.

32. The system according to claim 31, wherein the processing means further includes a context determining means that determines context of the utterance by applying prior probabilities or fuzzy possibilities to keyword matching, user profile or dialog history.

33. The system according to claim 32, wherein the domain is determined at least in part by the context.