Method and apparatus for updating speech recognition databases and reindexing audio and video content using the same

• US 20070106685A1
• Filed: 09/18/2006
• Published: 05/10/2007
• Est. Priority Date: 11/09/2005
• Status: Abandoned Application

• Current Assignee: Ramp Holdings, Inc. (Massachusetts)
• Original Assignee: Ramp, Inc.
• Inventors: Houh, Henry, Stern, Jeffrey, Zinovieva, Nina, Meteer, Marie
• Application Number:
  US11/522,645
• Publication Number:
  US 20070106685A1
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G06F 16/23 Updating
• CPC Class Codes: G06F 16/25 Integrating or interfacing systems involving database management systems
• CPC Class Codes: G06F 16/41 Indexing; Data structures therefor; Storage structures
• CPC Class Codes: G06F 16/43 Querying
• CPC Class Codes: G06F 16/438 Presentation of query results
• CPC Class Codes: G06F 16/483 using metadata automatically derived from the content
• CPC Class Codes: G06F 16/738 Presentation of query results
• CPC Class Codes: G06F 16/7844 using original textual content or text extracted from visual content or transcript of audio data

Abstract

A method and apparatus for reindexing media content for search applications that includes steps and structure for providing a speech recognition database that include entries defining acoustical representations for a plurality of words; providing a searchable database containing a plurality of metadata documents descriptive of a plurality of media resources, each of the plurality of metadata documents including a sequence of speech recognized text indexed using the speech recognition database; updating the speech recognition database with at least one word candidate; and reindexing the sequence of speech recognized text for a subset of the plurality of metadata documents using the updated speech recognition database.

Claims

1. A method for reindexing media content for search applications, comprising:

providing a speech recognition database that include entries defining acoustical representations for a plurality of words;

providing a searchable database containing a plurality of metadata documents descriptive of a plurality of media resources, each of the plurality of metadata documents including a sequence of speech recognized text indexed using the speech recognition database;

updating the speech recognition database with at least one word candidate; and

reindexing the sequence of speech recognized text for a subset of the plurality of metadata documents using the updated speech recognition database, the subset of metadata documents including metadata documents having a sequence of speech recognized text generated before the speech recognition database was updated with the at least one word candidate.

2. The method of claim 1 further comprising:

obtaining the at least one word candidate from one or more sources; and

reindexing the sequence of speech recognized text for a subset of the plurality of metadata documents using the updated speech recognition database, the subset of metadata documents including metadata documents having a sequence of speech recognized text generated before the at least one word candidate was obtained from the one or more sources.

3. The method of claim 1 further comprising:

scheduling a media resource for reindexing using the updated speech recognition database with a high priority if the content of the media resource and the at least one word candidate are associated with a common category.

4. The method of claim 1 further comprising:

scheduling the media resource for reindexing using the updated speech recognition database with a low priority if the content of the media resource and the at least one word candidate are associated with different categories.

5. The method of claim 1 wherein updating the speech recognition database with at least one word includes adding an entry to the speech recognition database that maps the at least one word candidate to an acoustical representation.

6. The method of claim 5 wherein the entry is added to a dictionary of the speech recognition database.

7. The method of claim 5 wherein the entry is added to a language model of the speech recognition database.

8. The method of claim 1 wherein updating the speech recognition database with at least one word includes adding a rule to a post-processing rules database, the rule defining criteria for replacing one or more words in a sequence of speech recognized text with the at least one word candidate during a post processing step.

9. The method of claim 1 wherein each of the acoustical representations is a string of phonemes.

10. The method of claim 1 wherein the plurality of words includes individual words or multiple word strings.

11. The method of claim 1, further comprising:

obtaining metadata descriptive of a media resource, the metadata comprising a first address to a first web site that provides access to the media resource;

accessing the first web site using the first address to obtain data from the web site; and

selecting the at least one word candidate from the text of words collected or derived from the data obtained from the first web site; and

updating the speech recognition database with the at least one word candidate.

12. The method of claim 11 wherein the at least one word candidate includes one or more frequently occurring words from the data obtained from the first web site.

13. The method of claim 11 further comprising:

accessing the first web site to identify one or more related web sites, the related web sites being linked to or referenced by the first web site;

obtaining web page data from the one or more related web sites;

selecting the at least one word candidate from the text of words collected or derived from the web page data obtained from the related web sites; and

updating the speech recognition database with the at least one word candidate.

14. The method of claim 1 further comprising:

obtaining metadata descriptive of a media resource, the metadata includes descriptive text of the media resource;

selecting the at least one word candidate from the descriptive text of the metadata; and

updating the speech recognition database with the at least one word candidate.

15. The method of claim 14 wherein the descriptive text of the metadata comprises a title, description or a link to the media resource.

16. The method of claim 14 wherein the descriptive text of the metadata comprises information from a web page describing the media resource.

17. The method of claim 1, further comprising:

obtaining web page data from a selected set of web sites;

selecting the at least one word candidate from the text of words collected or derived from the web page data obtained from the related web sites; and

updating the speech recognition database with the at least one word candidate.

18. The method of claim 17 wherein the at least one word candidate includes one or more frequently occurring words from the data obtained from the selected set of web sites.

19. The method of claim 1 further comprising:

tracking a plurality of search requests received by a search engine, each search request including one or more search query terms; and

selecting the at least one word candidate from the one or more search query terms.

20. The method of claim 19 wherein the at least one word candidate includes one or more search terms comprising a set of topmost requested search terms.

21. The method of claim 1 further comprising:

generating an acoustical representation for the at least one word candidate, the acoustical representation being associated with a confidence score; and

updating the speech recognition database with the at least one word candidate, the at least one word having a confidence score that satisfies a predetermined threshold.

22. The method of claim 21 further comprising:

excluding the at least one word candidate from the speech recognition database, the at least one word having a confidence score that fails to satisfy a predetermined threshold.

23. The method of claim 1 wherein the plurality of media resources comprising an audio resource or a video resource.

24. The method of claim 23 wherein the plurality of media resource comprises an audio or video podcast.

25. The method of claim 1 wherein reindexing the sequence of speech recognized text comprises reindexing less than all of the speech recognized text.

26. The method of claim 1 wherein reindexing the sequence of speech recognized text comprises reindexing all of the speech recognized text.

27. The method of claim 1 further comprising:

scheduling a media resource for partial reindexing using the updated speech recognition database if the metadata document corresponding to the media resource contains one or more phonetically similar words to the at least one word candidate added to the speech recognition database.

28. The method of claim 1 wherein a metadata document further comprises a sequence of phonemes derived from a media resource further comprising:

scheduling the media resource for partial reindexing using the updated speech recognition database if the metadata document contains at least one phonetically similar region to the constituent phonemes of the at least one word candidate added to the speech recognition database.

29. An apparatus for reindexing media content for search applications, comprising:

a speech recognition database that includes entries defining acoustical representations for a plurality of words;

a searchable database containing a plurality of metadata documents descriptive of a plurality of media resources,

a media indexer that generates a sequence of speech recognized text included in each of the plurality of metadata documents using the speech recognition database;

an update module that updates the speech recognition database with at least one word candidate; and

a reindexing module that causes the media indexer to reindex the sequence of speech recognized text for a subset of the plurality of metadata documents using the updated speech recognition database, the subset of metadata documents including metadata documents having a sequence of speech recognized text generated before the speech recognition database was updated with the at least one word candidate.

30. An apparatus for reindexing media content for search applications, comprising:

means for providing a speech recognition database that include entries defining acoustical representations for a plurality of words;

means for providing a searchable database containing a plurality of metadata documents descriptive of a plurality of media resources, each of the plurality of metadata documents including a sequence of speech recognized text indexed using the speech recognition database;

means for updating the speech recognition database with at least one word candidate; and

means for reindexing the sequence of speech recognized text for a subset of the plurality of metadata documents using the updated speech recognition database, the subset of metadata documents including metadata documents having a sequence of speech recognized text generated before the speech recognition database was updated with the at least one word candidate.