Method and system for classifying media content

• US 20070112838A1
• Filed: 06/07/2006
• Published: 05/17/2007
• Est. Priority Date: 06/07/2005
• Status: Active Grant

• Current Assignee: Getty Images Incorporated
• Original Assignee: Getty Images Incorporated
• Inventors: Gandert, Nate, Heck, Steven, O'Leary, Thomas, Bjarnestam, Anna, MacGuffie, Monika, Forster, Mary, Nance, Alvin, Huang, Ziji
• Granted Patent: US 10,445,359 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G06F 16/48 Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually

Abstract

A hardware and software facility for classifying media content units using keywords from a structured vocabulary. Metadata associated with each media content unit is segmented into a series of descriptive phrases. The descriptive phrases are mapped to keywords in a structured vocabulary, and the identified keywords associated with the media content units. Descriptive phrases that are not found in the structured vocabulary are tracked as candidate phrases for later addition to the structured vocabulary. A keyword index to the media content units may be constructed. The index is used to identify specific media content units that are responsive to search queries in a reliable and accurate fashion.

Claims

1. A method of processing metadata associated with media content units in order to classify the media content units using terms contained in a structured vocabulary, the method comprising:

(a) receiving metadata for association with a media content unit;

(b) programmatically segmenting the received metadata to generate one or more descriptive phrases;

(c) comparing the one or more descriptive phrases with a structured vocabulary of terms in order to identify one or more terms in the structured vocabulary that are related to the one or more descriptive phrases; and

(d) creating an association between the identified one or more terms and the media content unit so that the media content unit is characterized by the one or more terms.

2. The method of claim 1, further comprising the step of processing the metadata to remove a portion of the metadata prior to segmenting the metadata.

3. The method of claim 2, wherein processing the metadata includes removing numbers from the metadata.

4. The method of claim 2, wherein processing the metadata includes removing parentheticals from the metadata.

5. The method of claim 2, wherein processing the metadata includes removing source attributions from the metadata.

6. The method of claim 2, wherein processing the metadata includes removing non-grammatical symbols from the metadata.

7. The method of claim 2, wherein processing the metadata includes removing structured metadata from the metadata.

8. The method of claim 1, wherein the metadata is segmented using statistical processing.

9. The method of claim 1, wherein the metadata is segmented using linguistic processing.

10. The method of claim 1, wherein the metadata is segmented using an application-specific dictionary.

11. The method of claim 10, wherein the application-specific dictionary is a subset of the terms in the structured vocabulary of terms

12. The method of claim 1, wherein the media content unit is visual content.

13. The method of claim 12, wherein the visual content is moving imagery.

14. The method of claim 12, wherein the visual content is an image.

15. The method of claim 1, wherein comparing the one or more descriptive phrases with a structured vocabulary of terms includes identifying one or more descriptive phrases that are not in the structured vocabulary.

16. The method of claim 15, wherein the one or more descriptive phrases that are not in the structured vocabulary are maintained in a list of potential phrases to add to the structured vocabulary.

17. A system for processing metadata associated with media content units in order to classify the media content units using terms contained in a structured vocabulary, the system comprising:

(a) a segmentation component for receiving metadata associated with a plurality of media content units and segmenting the metadata to generate one or more descriptive phrases characterizing the plurality of media content units;

(b) a matching component for comparing the one or more descriptive phrases generated by the segmentation component with a structured vocabulary of terms in order to identify one or more terms in the structured vocabulary that are correlated with the one or more descriptive phrases for the plurality of media content units; and

(c) a mapping component that, for the plurality of media content units, provides a relationship between the media content unit and the one or more terms in the structured vocabulary that were identified by the matching component.

18. The system of claim 17, wherein the segmentation component further processes the metadata to remove a portion of the metadata prior to segmenting the metadata.

19. The system of claim 18, wherein processing the metadata includes removing numbers from the metadata.

20. The system of claim 18, wherein processing the metadata includes removing parentheticals from the metadata.

21. The system of claim 18, wherein processing the metadata includes removing source attributions from the metadata.

22. The system of claim 18, wherein processing the metadata includes removing non-grammatical symbols from the metadata.

23. The system of claim 18, wherein processing the metadata includes removing structured metadata from the metadata.

24. The system of claim 17, wherein the segmentation component segments the metadata using statistical processing.

25. The system of claim 17, wherein the segmentation component segments the metadata using linguistic processing.

26. The system of claim 17, wherein the segmentation component segments the metadata using an application-specific dictionary.

27. The system of claim 26, wherein the application-specific dictionary is a subset of the terms in the structured vocabulary of terms.

28. The system of claim 17, wherein the media content unit is visual content.

29. The system of claim 28, wherein the visual content is moving imagery.

30. The system of claim 28, wherein the visual content is an image.

31. The system of claim 17, wherein the matching component further identifies one or more descriptive phrases that are not in the structured vocabulary of terms.

32. The system of claim 31, wherein the matching component further adds the identified one or more descriptive phrases that are not in the structured vocabulary to a list of potential phrases to add to the structured vocabulary.

33. A method of processing descriptive text associated with visual content in order to classify the visual content using terms contained in a structured vocabulary, the method comprising:

(a) receiving descriptive text associated with visual content;

(b) programmatically segmenting the received descriptive text to identify one or more descriptive phrases;

(c) comparing the identified one or more descriptive phrases with a structured vocabulary of terms in order to identify one or more terms in the structured vocabulary that are related to the one or more descriptive phrases; and

(d) creating an association between the identified one or more terms and the image so that the visual content is classified by the one or more terms.

34. The method of claim 33, further comprising the step of processing the descriptive text to remove a portion of the descriptive text prior to segmenting the descriptive text.

35. The method of claim 34, wherein processing the descriptive text includes removing numbers from the descriptive text.

36. The method of claim 34, wherein processing the descriptive text includes removing parentheticals from the descriptive text.

37. The method of claim 34, wherein processing the descriptive text includes removing source attributions from the descriptive text.

38. The method of claim 34, wherein processing the metadata includes removing non-grammatical symbols from the metadata.

39. The method of claim 34, wherein processing the metadata includes removing structured metadata from the metadata

40. The method of claim 33, wherein the descriptive text is segmented using statistical processing.

41. The method of claim 33, wherein the descriptive text is segmented using linguistic processing.

42. The method of claim 33, wherein the descriptive text is segmented using an application-specific dictionary.

43. The method of claim 42, wherein the application-specific dictionary is a subset of the terms in the structured vocabulary of terms.

44. The method of claim 33, wherein comparing the one or more descriptive phrases with a structured vocabulary of terms includes identifying one or more descriptive phrases that are not in the structured vocabulary.

45. The method of claim 44, wherein the one or more descriptive phrases that are not in the structured vocabulary are maintained in a list of potential phrases to add to the structured vocabulary.

46. The method of claim 33, wherein the visual content is moving imagery.

47. The method of claim 33, wherein the visual content is an image.

48. A computer-readable medium containing instructions for controlling a computer processor in a computer system to classify media content units using terms contained in a structured vocabulary by:

(a) receiving metadata associated with a media content unit;

(b) programmatically segmenting the received metadata to generate one or more descriptive phrases;

(c) comparing the one or more descriptive phrases with a structured vocabulary of terms in order to identify one or more terms in the structured vocabulary that are related to the one or more descriptive phrases; and

(d) creating an association between the identified one or more terms and the media content unit so that the media content unit is classified by the one or more terms.

49. The computer-readable medium of claim 48, further comprising processing the metadata to remove a portion of the metadata prior to segmenting the metadata.

50. The computer-readable medium of claim 48, wherein the metadata is segmented using an application-specific dictionary.

51. The method of claim 50, wherein the application-specific dictionary is a subset of the terms in the structured vocabulary of terms.

52. A system for segmenting metadata associated with a plurality of media content units, the system comprising:

a structured keyword vocabulary used to classify media content units; and

a segmentation component for receiving metadata for association with a plurality of media content units and segmenting the metadata to generate one or more descriptive phrases characterizing the plurality of media content units, wherein the segmentation component utilizes an application-specific dictionary that is a subset of the structured keyword vocabulary to segment the received metadata.

53. The system of claim 52, wherein the application-specific dictionary excludes keywords in the structured keyword vocabulary that primarily relate to editorial functions.

54. A method of segmenting metadata associated with a plurality of media content units, the method comprising:

(a) receiving metadata for association with a plurality of media content units; and

(b) programmatically segmenting the received metadata to generate one or more descriptive phrases associated with the plurality of media content units utilizing an application-specific dictionary, wherein the application-specific dictionary is a subset of a structured keyword vocabulary that is used to classify the plurality of media content units.

55. The method of claim 54, wherein the application-specific dictionary excludes keywords in the structured vocabulary that primarily relate to editorial functions.