Resolving pronoun ambiguity in voice queries

• US 9,529,793 B1
• Filed: 02/22/2013
• Issued: 12/27/2016
• Est. Priority Date: 06/01/2012
• Status: Active Grant

• Current Assignee: Google LLC (Alphabet Inc.)
• Original Assignee: Google Inc. (Alphabet Inc.)
• Inventors: Taubman, Gabriel, Lee, John J.
• Primary Examiner(s): PATEL, SHREYANS A
• Application Number:
  US13/774,082
• Time in Patent Office: 1,404 Days
• Field of Search: 704/9
• US Class Current: 1/1
• CPC Class Codes: G06F 16/635 Filtering based on additional data, e.g. user or group profiles
• CPC Class Codes: G06F 16/951 Indexing; Web crawling techniques
• CPC Class Codes: G06F 40/253 Grammatical analysis; Style critique
• CPC Class Codes: G10L 15/187 Phonemic context, e.g. pronunciation rules, phonotactical constraints or phoneme n-grams
• CPC Class Codes: G10L 25/87 Detection of discrete points within a voice signal

Abstract

Methods, systems, and apparatus, including computer programs encoded on computer storage media, for resolving ambiguity in received voice queries. An original voice query is received following one or more earlier voice queries, wherein the original voice query includes a pronoun or phrase. In one implementation, a plurality of acoustic parameters is identified for one or more words in the original voice query. A concept represented by the pronoun is identified based on the plurality of acoustic parameters, wherein the concept is associated with a particular query of the one or more earlier queries. The concept is associated with the pronoun. Alternatively, a concept may be associated with a phrase by using grammatical analysis of the query to relate the phrase to a concept derived from a prior query.

Claims

1. A computer-implemented method, comprising:

receiving a voice query that includes a pronoun;

determining a stress score for the pronoun that reflects an extent to which the pronoun was stressed in the voice query;

identifying two or more candidate entities that are associated with two or more answers that a search engine generates for two or more corresponding voice queries that preceded the voice query;

selecting, as a particular entity, (i) an entity associated with an answer that the search engine generates for an immediately preceding voice query when the stress score satisfies a predetermined threshold, or (ii) an entity associated with an answer that is generated for a voice query that is not the immediately preceding voice query when the stress score does not satisfy the predetermined threshold;

revising the received voice query to refer to the particular entity that was selected based at least on the comparison of the stress score to the predetermined threshold; and

providing one or more query results that are responsive to the revised voice query for output on a client device.

2. The method of claim 1, wherein the determined stress score is based on a plurality of acoustic parameters associated with the voice query.

3. The method of claim 2, wherein the acoustic parameters include two or more of volume, pitch, frequency, duration between each spoken word, and spoken duration.

10. The method of claim 1, further comprising:

identifying a first set of acoustic parameters that includes one or more acoustic parameters associated with the pronoun in the voice query;

identifying a second set of acoustic parameters that includes one or more acoustic parameters associated with one or more other words in the voice query; and

determining a stress score for the pronoun that reflects an extent to which the pronoun was stressed in the voice query, based at least on a comparison of the first set of acoustic parameters and the second set of acoustic parameters.

11. The method of claim 1, further comprising:

associating a confidence score with the stress score; and

selecting a particular entity, from among the two or more candidate entities that are associated with the two or more answers that a search engine generates for two or more corresponding voice queries that preceded the voice query, based at least on the stress score and the confidence score.

12. The method of claim 1, further comprising:

determining a gender score based on the gender that the pronoun in the voice query references; and

selecting a particular entity, from among the two or more candidate entities that are associated with two or more answers that a search engine generates for two or more corresponding voice queries that preceded the voice query, based at least on the stress score and the gender score.

4. A system comprising:

one or more computers and one or more storage devices storing instructions that are operable, when executed by the one or more computers, to cause the one or more computes to perform operations comprising;

receiving a voice query that includes a pronoun;

determining a stress score for the pronoun that reflects an extent to which the pronoun was stressed in the voice query;

identifying two or more candidate entities that are associated with two or more answers that a search engine generates for two or more corresponding voice queries that preceded the voice query;

selecting, as a particular entity, (i) an entity associated with an answer that the search engine generates for an immediately preceding voice query when the stress score satisfies a predetermined threshold, or (ii) an entity associated with an answer that is generated for a voice query that is not the immediately preceding voice query when the stress score does not satisfy the predetermined threshold;

revising the received voice query to refer to the particular entity that was selected based on the comparison of the stress score to the predetermined threshold; and

providing one or more query results that are responsive to the revised voice query for output on a client device.

5. The system of claim 4, wherein the determined stress score is based on a plurality of acoustic parameters associated with the voice query.

6. The system of claim 5, wherein the acoustic parameters include two or more of volume, pitch, frequency, duration between each spoken word, and spoken duration.

13. The system of claim 4, wherein the operations further comprise:

identifying a first set of acoustic parameters that includes one or more acoustic parameters associated with the pronoun in the voice query;

identifying a second set of acoustic parameters that includes one or more acoustic parameters associated with one or more other words in the voice query; and

determining a stress score for the pronoun that reflects an extent to which the pronoun was stressed in the voice query, based at least on a comparison of the first set of acoustic parameters and the second set of acoustic parameters.

14. The system of claim 4, wherein the operations further comprise:

associating a confidence score with the stress score; and

selecting a particular entity, from among the two or more candidate entities that are associated with the two or more answers that a search engine generates for two or more corresponding voice queries that preceded the voice query, based at least on the stress score and the confidence score.

15. The system of claim 4, wherein the operations further comprise:

determining a gender score based on the gender that the pronoun in the voice query references; and

selecting a particular entity, from among the two or more candidate entities that are associated with two or more answers that a search engine generates for two or more corresponding voice queries that preceded the voice query, based at least on the stress score and the gender score.

7. A non-transitory computer-readable medium storing software comprising instructions executable by one or more computers which, upon such execution, cause the one or more computers to perform operations comprising:

receiving a voice query that includes a pronoun;

determining a stress score for the pronoun that reflects an extent to which the pronoun was stressed in the voice query;

identifying two or more candidate entities that are associated with two or more answers that a search engine generates for two or more corresponding voice queries that preceded the voice query;

selecting, as a particular entity, (i) an entity associated with an answer that the search engine generates for an immediately preceding voice query when the stress score satisfies a predetermined threshold, or (ii) an entity associated with an answer that is generated for a voice query that is not the immediately preceding voice query when the stress score does not satisfy the predetermined threshold; and

revising the received voice query to refer to the particular entity that was selected based on the comparison of the stress score to the predetermined threshold; and

providing one or more query results that are responsive to the revised voice query for output on a client device.

8. The medium of claim 7, wherein the determined stress score is based on a plurality of acoustic parameters associated with the voice query.

9. The medium of claim 8, wherein the acoustic parameters include two or more of volume, pitch, frequency, duration between each spoken word, and spoken duration.

16. The medium of claim 7, wherein the operations further comprise:

identifying a first set of acoustic parameters that includes one or more acoustic parameters associated with the pronoun in the voice query;

identifying a second set of acoustic parameters that includes one or more acoustic parameters associated with one or more other words in the voice query; and

determining a stress score for the pronoun that reflects an extent to which the pronoun was stressed in the voice query, based at least on a comparison of the first set of acoustic parameters and the second set of acoustic parameters.

17. The medium of claim 7, wherein the operations further comprise:

associating a confidence score with the stress score; and

selecting a particular entity, from among the two or more candidate entities that are associated with the two or more answers that a search engine generates for two or more corresponding voice queries that preceded the voice query, based at least on the stress score and the confidence score.

18. The medium of claim 7, wherein the operations further comprise:

determining a gender score based on the gender that the pronoun in the voice query references; and

selecting a particular entity, from among the two or more candidate entities that are associated with two or more answers that a search engine generates for two or more corresponding voice queries that preceded the voice query, based at least on the stress score and the gender score.