Auto-Generation of Parsing Grammars from a Concept Ontology

• US 20140297282A1
• Filed: 03/28/2013
• Published: 10/02/2014
• Est. Priority Date: 03/28/2013
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Nuance Communications, Inc. (Microsoft Corporation)
• Inventors: Peters, Stephen Douglas, Tremblay, Ral
• Granted Patent: US 10,811,004 B2
• Time in Patent Office: Days
• US Class Current: 704/254
• CPC Class Codes: G10L 15/1815 Semantic context, e.g. disambiguation of the recognition hypotheses based on word meaning
• CPC Class Codes: G10L 15/19 Grammatical context, e.g. disambiguation of the recognition hypotheses based on word sequence rules

Abstract

An ontology stores information about a domain of an automatic speech recognition (ASR) application program. The ontology is augmented with information that enables subsequent automatic generation of a speech understanding grammar for use by the ASR application program. The information includes hints about how a human might talk about objects in the domain, such as preludes (phrases that introduce an identification of the object) and postludes (phrases that follow an identification of the object).

Claims

1. A computerized ontology for facilitating automatically generating a speech understanding grammar for an automatic speech recognition (ASR) application program, the ontology comprising:

a computerized database containing information about a plurality of hierarchically organized concepts related to a domain of the ASR application program; and

for each of at least a subset of the plurality of concepts, information associated with the concept, wherein the information comprises at least one hint about how a human speaker might refer to the concept in an utterance.

2. A computerized ontology according to claim 1, wherein the at least one hint comprises a prelude and a postlude.

3. A computer-implemented method for automatically generating a speech understanding grammar for an automatic speech recognition (ASR) application program, the method comprising performing by a processor operations of:

accessing a computerized ontology that contains;

information about a plurality of hierarchically organized concepts related to a domain of the ASR application program; and

for each concept of at least a subset of the plurality of concepts, at least one associated hint about how a human speaker might refer to the concept in an utterance;

accepting, via a user interface, at least one user command that identifies a user-selected subset of the plurality of concepts;

retrieving from the ontology at least one hint associated with each concept of the user-selected subset of the plurality of concepts; and

generating a speech understanding grammar from the retrieved at least one hint.

4. A method according to claim 3, further comprising, for at least one concept of the user-selected subset of the plurality of concepts:

identifying a second concept at a lower hierarchical level than the at least one concept; and

retrieving from the ontology at least one second hint associated with the second concept;

wherein;

generating the speech understanding grammar from the retrieved at least one hint comprises generating a speech understanding grammar from the retrieved at least one second hint.

5. A method according to claim 3, wherein the at least one hint comprises a prelude and a postlude.

6. A method for constructing a database to facilitate subsequently automatically generating a speech understanding grammar for an automatic speech recognition (ASR) application program, the method comprising:

in a computerized ontology that contains information about a plurality of hierarchically organized concepts related to a domain of the ASR application program, for each of at least a subset of the plurality of concepts, storing information in association with the concept, wherein the information comprises at least one hint about how a human speaker might refer to the concept in an utterance.

7. A method according to claim 6, wherein the at least one hint comprises a prelude and a postlude.

8. A computer program product for providing automatically generating a speech understanding grammar for an automatic speech recognition (ASR) application program, the computer program product comprising a non-transitory computer-readable medium having computer readable program code stored thereon, the computer readable program code configured to:

access a computerized ontology that contains;

information about a plurality of hierarchically organized concepts related to a domain of the ASR application program; and

for each concept of at least a subset of the plurality of concepts, at least one associated hint about how a human speaker might refer to the concept in an utterance;

accept, via a user interface, at least one user command that identifies a user-selected subset of the plurality of concepts;

retrieve from the ontology at least one hint associated with each concept of the user-selected subset of the plurality of concepts; and

generate a speech understanding grammar from the retrieved at least one hint.

9. A computer program product according to claim 8, wherein the computer readable program code is further configured to, for at least one concept of the user-selected subset of the plurality of concepts:

identify a second concept at a lower hierarchical level than the at least one concept;

retrieve from the ontology at least one second hint associated with the second concept; and

generate the speech understanding grammar from the retrieved at least one hint by generating a speech understanding grammar from the retrieved at least one second hint.

10. A computer program product according to claim 8, wherein the at least one hint comprises a prelude and a postlude.