System and Method for Relating Syntax and Semantics for a Conversational Speech Application

• US 20070265847A1
• Filed: 06/07/2007
• Published: 11/15/2007
• Est. Priority Date: 01/12/2001
• Status: Active Grant

• Current Assignee: Nuance Communications, Inc. (Microsoft Corporation)
• Original Assignee: Nuance Communications, Inc. (Microsoft Corporation)
• Inventors: Ross, Steven, MacAllister, Jeffrey, Brownholtz, Elizabeth, Alweis, Julie, Armes, Robert
• Granted Patent: US 8,438,031 B2
• Time in Patent Office: Days
• US Class Current: 704/236
• 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 2015/228 of application context

Abstract

A conversation manager processes spoken utterances from a user of a computer. The conversation manager includes a semantics analysis module and a syntax manager. A domain model that is used in processing the spoken utterances includes an ontology (i.e., world view for the relevant domain of the spoken utterances), lexicon, and syntax definitions. The syntax manager combines the ontology, lexicon, and syntax definitions to generate a grammatic specification. The semantics module uses the grammatic specification and the domain model to develop a set of frames (i.e., internal representation of the spoken utterance). The semantics module then develops a set of propositions from the set of frames. The conversation manager then uses the set of propositions in further processing to provide a reply to the spoken utterance.

Claims

1. A computer method of generating a grammar for analyzing spoken utterances in a speech enabled application interface, comprising:

providing a domain model of propositions for interfacing with a software application, the domain model including an ontology, lexicon and syntax template;

through the syntax template, applying a general syntax to an ontological model such that an annotated grammar is formed; and

generating a grammatic specification based on the domain model and including the annotated grammar, the grammatic specification stored to a computer-readable storage medium and suitable for processing spoken utterances directed to interfacing with the software application.

2. The computer method of claim 1, wherein the ontology includes an ontological description of the domain model based on entities, classes, and attributes.

3. The computer method of claim 1, wherein the syntax template specifies legal word sequences based on the ontological description.

4. The computer method of claim 1, wherein the generating the grammatic specification includes combining the ontological description, lexicon, and the syntax templates to generate the grammatic specification.

5. The computer method of claim 1, wherein the lexicon provides synonyms and parts of speech information for elements of the ontological description, and wherein the grammatic specification is based on the lexicon.

6. The computer method of claim 1, wherein the domain model is based on an application-independent domain model.

7. The computer method of claim 6, wherein the domain model includes an application domain model having element corresponding to propositions specific to interfacing with the speech-enabled application.

8. A computer program product comprising:

a computer usable medium for analyzing spoken utterances comprising common language words in a speech-enabled environment; and

a set of computer program instructions embodied on the computer usable medium, including instructions to;

provide a domain model of propositions for interfacing with a software application, the domain model including an ontology, lexicon and syntax template;

through the syntax template, apply a general syntax to an ontological model such that an annotated grammar is formed; and

generate a grammatic specification based on the domain model and including the annotated grammar, the grammatic specification stored to a computer-readable storage medium and suitable for processing spoken utterances directed to interfacing with the software application.

9. The computer program product of claim 8, wherein the ontology includes an ontological description of the domain model based on entities, classes, and attributes.

10. The computer program product of claim 8, wherein the syntax template specifies legal word sequences based on the ontological description.

11. The computer program product of claim 8, wherein the generating the grammatic specification includes combining the ontological description, lexicon, and the syntax templates to generate the grammatic specification.

12. The computer program product of claim 8, wherein the lexicon provides synonyms and parts of speech information for elements of the ontological description, and wherein the grammatic specification is based on the lexicon.

13. The computer program product of claim 8, wherein the domain model is based on an application-independent domain model.

14. The computer program product of claim 13, wherein the domain model includes an application domain model having element corresponding to propositions specific to interfacing with the speech-enabled application.

15. A computer method of generating a grammar for analyzing spoken utterances in a speech enabled application interface, comprising:

receiving an ontological description of a domain model based on entities, classes and attributes, the domain model specific to a speech enabled application;

receiving syntax templates for the domain model specifying legal word sequences based on the ontological description; and

combining the ontological description, a lexicon, and the syntax templates to generate a grammatic specification, the grammatic specification stored to a computer readable storage medium and suitable for processing spoken utterances for interfacing with the speech enabled application.

16. The computer method of claim 15, wherein the lexicon provides synonyms and parts of speech information for elements of the ontological description, and wherein the grammatic specification is based on the lexicon.

17. The computer method of claim 15, wherein the ontology includes an ontological description of the domain model based on entities, classes, and attributes.

18. The computer method of claim 15, wherein the syntax template specifyies legal word sequences based on the ontological description.

19. The computer method of claim 15, wherein the generating the grammatic specification includes combining the ontological description, lexicon, and the syntax templates to generate the grammatic specification.

20. A computer method of providing a speech-enabled interface, comprising:

detecting a spoken utterance directed to a speech-enabled application;

processing the spoken utterance using a grammatic specification based on an domain model for the application to produce an initial semantic representation based on the recognized spoken utterance;

providing a set of propositions that represent the recognized spoken utterance based on the initial semantic representation;

determining a goal corresponding to the set of propositions; and

generating a response to the set of propositions according to the goal.

21. The computer method of claim 20, wherein generating the response includes at least one of:

providing instructions to operate the speech-enabled application;

providing, to a user, information responsive to the set of propositions; and

providing an inquiry for information required to achieve the goal.

22. The computer method of claim 20, wherein the domain model includes an ontology, lexicon and syntax template.

23. The computer method of claim 22, wherein the ontology includes an ontological description of the domain model based on entities, classes, and attributes.

24. The computer method of claim 22, wherein the syntax template specifyies legal word sequences based on the ontological description.

25. The computer method of claim 22, wherein processing the spoken utterance includes analyzing the spoken utterence based on the grammatic specification, the grammatic specification incorporating the ontological description, lexicon, and the syntax template.

26. The computer method of claim 20, wherein the domain model is based on an application-independent domain model.

27. The computer method of claim 26, wherein the domain model includes an application domain model having element corresponding to propositions specific to interfacing with the speech-enabled application.

28. A computer program product comprising:

a computer usable medium for analyzing spoken utterances comprising common language words in a speech-enabled environment; and

a set of computer program instructions embodied on the computer usable medium, including instructions to;

detect a spoken utterance directed to a speech-enabled application;

process the spoken utterance using a grammatic specification based on an domain model for the application to produce an initial semantic representation based on the recognized spoken utterance;

provide a set of propositions that represent the recognized spoken utterance based on the initial semantic representation;

determine a goal corresponding to the set of propositions; and

generate a response to the set of propositions according to the goal.

29. The computer method of claim 28, wherein generating the response includes at least one of:

providing instructions to operate the speech-enabled application;

providing, to a user, information responsive to the set of propositions; and

providing an inquiry for information required to achieve the goal.

30. The computer method of claim 28, wherein the domain model includes an ontology, lexicon and syntax template.

31. The computer method of claim 28, wherein the ontology includes an ontological description of the domain model based on entities, classes, and attributes.

32. The computer method of claim 28, wherein the syntax template specifyies legal word sequences based on the ontological description.

33. The computer method of claim 28, wherein processing the spoken utterance includes analyzing the spoken utterence based on the grammatic specification, the grammatic specification incorporating the ontological description, lexicon, and the syntax template.

34. The computer method of claim 28, wherein the domain model is based on an application-independent domain model.

35. The computer method of claim 34, wherein the domain model includes an application domain model having element corresponding to propositions specific to interfacing with the speech-enabled application.