Navigation using a search engine and phonetic voice recognition

• US 8,380,512 B2
• Filed: 03/10/2008
• Issued: 02/19/2013
• Est. Priority Date: 03/10/2008
• Status: Active Grant

• Current Assignee: R2 Solutions LLC (Acacia Research Corporation)
• Original Assignee: Yahoo! Inc. (Apollo Global Management, Inc.)
• Inventors: Bonforte, Jeffrey, Clayton, Gary, Chen, Victor
• Primary Examiner(s): Godbold, Douglas
• Application Number:
  US12/045,365
• Publication Number:
  US 20090228277A1
• Time in Patent Office: 1,807 Days
• Field of Search: 704/270, 704/275, 704/1, 704/9
• US Class Current: 704/270
• CPC Class Codes: G06F 16/632 Query formulation
• CPC Class Codes: G10L 15/26 Speech to text systems G10L15/08 takes precedence
• CPC Class Codes: G10L 2015/025 Phonemes, fenemes or fenones being the recognition units

Abstract

A method and apparatus for incorporating voice recognition into a search engine is provided. The phonetic voice recognition system lacking grammar and spell checking is used. The output of the phonetic voice recognition system is forwarded to a search engine. The search engine performs disambiguation and relevancy analysis based on past similar queries. Search engine user behavior is recorded to improve the accuracy. Recorded statistics are used to rank results pages.

Claims

1. A computer-implemented method comprising the steps of:

receiving, at a search engine, a phonetically-spelled string;

wherein the phonetically-spelled string is the result of transforming an audio input into the phonetically-spelled string that has not been converted into any predetermined set of correctly-spelled words;

identifying one or more previously-submitted phonetically-spelled query strings from a plurality of other users based on the phonetically-spelled string; and

generating a set of query results based, at least in part, on the phonetically-spelled string and on the one or more previously-submitted phonetically-spelled query strings;

wherein the steps of the method are performed by one or more computing devices.

2. The method of claim 1 wherein transforming the audio input into the phonetically-spelled string is performed by a phonetic voice recognition component.

3. The method of claim 1 wherein transforming the audio input into the phonetically-spelled string includes disambiguating the phonetically-spelled string.

4. The method of claim 1 including a further step of recording one or more addresses of pages visited by a user, wherein each page was provided to the user in one or more previous search results.

5. The method of claim 4 wherein the set of query results is ordered, at least in part, using the recorded addresses.

6. The method of claim 1 including a further step of prompting a user for an indication of which query result of the set of query results is a best result.

7. The method of claim 6 wherein the step of prompting the user includes:

transforming a part of a selected result, from the set of results, from text to speech; and

sending an audio prompt to the user, wherein the audio prompt includes audio that corresponds to the part of the selected result.

8. The method of claim 1 wherein the set of query results is ordered based, at least in part, on past strings recorded by the search engine.

9. A non-transitory computer-readable storage medium storing one or more sequences of instructions, wherein execution of the one or more sequences of instructions by one or more processors causes the one or more processors to perform the steps of:

receiving, at a search engine, a phonetically-spelled string;

wherein the phonetically-spelled string is the result of transforming an audio input into the phonetically-spelled string that has not been converted into any predetermined set of correctly-spelled words;

identifying one or more previously-submitted phonetically-spelled query strings from a plurality of other users based on the phonetically-spelled string; and

generating a set of query results based, at least in part, on the phonetically-spelled string and on the one ore more previously-submitted phonetically-spelled query strings;

wherein the steps of the method are performed by one or more computing devices.

10. The non-transitory computer-readable storage medium of claim 9, wherein transforming the audio input into the phonetically-spelled string is performed by a phonetic voice recognition component.

11. The non-transitory computer-readable storage medium of claim 9, wherein transforming the audio input into the phonetically-spelled string includes a step of disambiguating the phonetically-spelled string.

12. The non-transitory computer-readable storage medium of claim 9, wherein execution of the one or more sequences of instructions by the one or more processors causes the one or more processors to further perform the step of:

recording one or more addresses of pages visited by a user, wherein each page was provided to the user in one or more previous search results.

13. The non-transitory computer-readable storage medium of claim 12, wherein the set of query results are ordered, at least in part, using the recorded addresses.

14. The non-transitory computer-readable storage medium of claim 9, wherein execution of the one or more sequences of instructions by the one or more processors causes the one or more processors to further perform the step of:

prompting a user for an indication of which query result of the set of query results is a best result.

15. The non-transitory computer-readable storage medium of claim 14, wherein prompting the user includes:

transforming a part of a selected result, from the set of results, from text to speech; and

sending an audio prompt to the user, wherein the audio prompt includes audio that corresponds to the part of the selected result.

16. The non-transitory computer-readable storage medium of claim 9, wherein the set of query results is ordered based, at least in part, on past strings recorded by the search engine.