Dictionary filtering using market data

• US 8,473,293 B1
• Filed: 10/23/2012
• Issued: 06/25/2013
• Est. Priority Date: 04/17/2012
• Status: Active Grant

• Current Assignee: Google LLC (Alphabet Inc.)
• Original Assignee: Google Inc. (Alphabet Inc.)
• Inventors: Sorensen, Jeffrey S., Mengibar, Pedro J.
• Primary Examiner(s): YEN, ERIC L
• Application Number:
  US13/658,139
• Time in Patent Office: 245 Days
• Field of Search: 704/10, 704243-244
• US Class Current: 704/243
• CPC Class Codes: G10L 15/06 Creation of reference templates; Training of speech recognition systems, e.g. adaptation to the characteristics of the speaker's voice G10L15/14 takes precedence

Abstract

This specification describes technologies relating to system, methods, and articles for updating a speech recognition dictionary based on, at least in part, both search query and market data metrics. In general, one innovative aspect of the subject matter described in this specification can be embodied in a method comprising (i) identifying a candidate term for possible inclusion in a speech recognition dictionary, (ii) identifying at least one search query metric associated with the identified candidate term, (iii) identifying at least one market data metric associated with the identified candidate term, and (iv) generating a candidate term score for the identified candidate term based, at least in part, on a weighted combination of the at least one identified search query metric and the at least one identified market data metric.

Claims

1. A system comprising:

one or more computers; and

a computer-readable medium coupled to the one or more computers having instructions stored thereon which, when executed by the one or more computers, cause the one or more computers to perform operations comprising;

identifying a candidate term for possible inclusion in a speech recognition dictionary;

identifying at least one search query metric associated with the identified candidate term;

identifying at least one market data metric associated with the identified candidate term; and

generating a candidate term score for the identified candidate term based, at least in part, on a weighted combination of the at least one identified search query metric and the at least one identified market data metric.

2. The system of claim 1, the operations further comprising:

identifying a candidate term score minimum threshold associated with the speech recognition dictionary; and

determining whether the generated candidate term score for the identified candidate term meets or exceeds the candidate term score minimum threshold.

3. The system of claim 2, where the identified candidate term is not included within the speech recognition dictionary, the operations further comprising:

adding the identified candidate term to the speech recognition dictionary in response to the determination that the generated candidate term score for the identified candidate term meets or exceeds the candidate term score minimum threshold.

4. The system of claim 3, the operations further comprising:

identifying a term previously included within the speech recognition dictionary corresponding to the relatively lowest candidate term score based on at least one search query metric associated with the previously included term and on at least one market data metric associated with the previously included term; and

removing the previously included term from the speech recognition dictionary in response to adding the identified candidate term to the speech recognition dictionary.

5. The system of claim 2, where the identified candidate term is not included within the speech recognition dictionary, the operations further comprising:

rejecting the identified candidate term from addition to the speech recognition dictionary in response to the determination that the generated candidate term score for the identified candidate term does not meet or exceed the candidate term score minimum threshold.

6. The system of claim 2, wherein the identified candidate term is previously included within the speech recognition dictionary, the operations further comprising:

removing the identified candidate term from the speech recognition dictionary in response to the determination that the generated candidate term score for the identified candidate term does not meet or exceed the candidate term score minimum threshold.

7. The system of claim 1, wherein the at least one search query metric includes at least one of the following:

a query count associated with a number of searches for the candidate term, click information associated with the candidate term, and a time-based search query metric identifying a query count within a predefined period of time.

8. The system of claim 1, wherein the at least one market data metric includes at least one of the following:

revenue generated in association with the candidate term, a count of advertising conversions associated with the candidate term, an advertising conversion rate associated with the candidate term, a number of impressions associated with the candidate term, an advertising bid price associated with the candidate term, a number of bids received for the candidate term, and a number of bidders associated with the candidate term.

9. The system of claim 1, wherein identifying the candidate term for possible inclusion in a speech recognition dictionary occurs in response to occurrence of a predetermined triggering update event.

10. A computer-implemented method performed by one or more processors, the method comprising:

identifying, by at least one processor, a candidate term for possible inclusion in a speech recognition dictionary;

identifying, by at least one processor, at least one search query metric associated with the identified candidate term;

identifying at least one market data metric associated with the identified candidate term; and

generating, by at least one processor, a candidate term score for the identified candidate term based, at least in part, on a weighted combination of the at least one identified search query metric and the at least one identified market data metric.

11. The method of claim 10, further comprising:

identifying a candidate term score minimum threshold associated with the speech recognition dictionary; and

determining whether the generated candidate term score for the identified candidate term meets or exceeds the candidate term score minimum threshold.

12. The method of claim 11, where the identified candidate term is not included within the speech recognition dictionary, the operations further comprising:

adding the identified candidate term to the speech recognition dictionary in response to the determination that the generated candidate term score for the identified candidate term meets or exceeds the candidate term score minimum threshold.

13. The method of claim 12, further comprising:

identifying a term previously included within the speech recognition dictionary corresponding to the relatively lowest candidate term score based on at least one search query metric associated with the previously included term and on at least one market data metric associated with the previously included term; and

removing the previously included term from the speech recognition dictionary in response to adding the identified candidate term to the speech recognition dictionary.

14. The method of claim 11, where the identified candidate term is not included within the speech recognition dictionary, the operations further comprising:

rejecting the identified candidate term from addition to the speech recognition dictionary in response to the determination that the generated candidate term score for the identified candidate term does not meet or exceed the candidate term score minimum threshold.

15. The method of claim 11, wherein the identified candidate term is previously included within the speech recognition dictionary, the operations further comprising:

removing the identified candidate term from the speech recognition dictionary in response to the determination that the generated candidate term score for the identified candidate term does not meet or exceed the candidate term score minimum threshold.

16. The method of claim 10, wherein the at least one search query metric includes at least one of the following:

a query count associated with a number of searches for the candidate term, or click information associated with the candidate term, and a time-based search query metric identifying a query count within a predefined period of time.

17. The method of claim 10, wherein the at least one market data metric includes at least one of the following:

revenue generated in association with the candidate term, a count of advertising conversions associated with the candidate term, an advertising conversion rate associated with the candidate term, and a number of impressions associated with the candidate term, an advertising bid price associated with the candidate term, a number of bids received for the candidate term, and a number of bidders associated with the candidate term.

18. An article of manufacture comprising a computer-readable storage device, the computer readable storage device storing instructions that are operable, when executed by one or more computers, to cause the one or more computers to perform operations comprising:

identifying a candidate term for possible inclusion in a speech recognition dictionary;

identifying at least one search query metric associated with the identified candidate term;

identifying at least one market data metric associated with the identified candidate term; and

generating a candidate term score for the identified candidate term based, at least in part, on a weighted combination of the at least one identified search query metric and the at least one identified market data metric.

19. The article of claim 18, the operations further including:

identifying a candidate term score minimum threshold associated with the speech recognition dictionary; and

determining whether the generated candidate term score for the identified candidate term meets or exceeds the candidate term score minimum threshold.

20. The article of claim 19, where the identified candidate term is not included within the speech recognition dictionary, the operations further comprising:

adding the identified candidate term to the speech recognition dictionary in response to the determination that the generated candidate term score for the identified candidate term meets or exceeds the candidate term score minimum threshold.