METHOD AND SYSTEM FOR IMPROVING THE QUALITY OF DEEP METADATA ASSOCIATED WITH MEDIA CONTENT

• US 20090198700A1
• Filed: 02/01/2008
• Published: 08/06/2009
• Est. Priority Date: 02/01/2008
• Status: Active Grant

• Current Assignee: Tahoe Research Limited (f/k/a Learndale Limited) (Vector Capital Corporation)
• Original Assignee: RealNetworks, Inc.
• Inventors: Steffens, Bjorn, Friedenberger, Norman, Spechtler, Andreas, Ross, Alexander
• Granted Patent: US 7,840,581 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
• CPC Class Codes: G06F 16/68 Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually

Abstract

Methods and systems verify and control the quality of metadata associated with a media data file. The metadata may be used, for example, by media content recommendation systems. In one embodiment, a first attribute is selected from metadata associated with a media data file and an attribute dependency corresponding to the first attribute is determined. The metadata may be searched to determine whether it includes a second attribute that satisfies the dependency. If the dependency is not satisfied, the metadata is flagged for reevaluation. The metadata may also be flagged for reevaluation, for example, if the metadata includes inconsistent attributes, if the total number of assigned attributes within the metadata does not exceed a predetermined value, or if an expected attribute dimension is missing or does not include an expected number of defined attributes. In certain embodiments, user selections are controlled to provide consistent attribute combinations.

Claims

1. A method for verifying the quality of metadata associated with a media data file, the method comprising:

selecting a first attribute from the metadata associated with the media data file;

determining an attribute dependency for the first attribute;

determining whether the metadata includes a second attribute that satisfies the dependency; and

if the dependency is not satisfied, flagging the metadata associated with the media data file for reevaluation.

2. The method of claim 1, further comprising:

selecting a third attribute from the metadata associated with the media data file;

determining whether the combination of the first attribute with the third attribute exceeds a probability threshold; and

if the combination does not exceed the probability threshold, flagging the metadata associated with the media data file for reevaluation.

3. The method of claim 1, further comprising:

determining whether a total number of attributes defined in the metadata associated with the media data file exceeds a predetermined value; and

if the total number of attributes does not exceed the predetermined value, flagging the metadata associated with the media data file for reevaluation.

4. The method of claim 1, further comprising:

searching the metadata associated with the media data file for an expected attribute dimension, the attribute dimension associated with a subset of attributes; and

if the expected attribute dimension is missing, flagging the metadata associated with the media data file for reevaluation.

5. The method of claim 4, further comprising:

determining whether the expected attribute dimension includes a threshold number of defined attributes in the subset; and

if the expected attribute dimension does not include the threshold number of defined attributes in the subset, flagging the metadata associated with the media data file for reevaluation.

6. The method of claim 1, further comprising:

searching the metadata associated with the media data file for an expected attribute; and

if the expected attribute is missing, flagging the metadata associated with the media data file for reevaluation.

7. A computer accessible medium comprising program instructions for causing a computer to perform a method for verifying the quality of metadata associated with a media data file, the method comprising:

selecting a first attribute from the metadata associated with the media data file;

selecting a second attribute from the metadata associated with the media data file;

determining whether the combination of the first attribute with the second attribute exceeds a probability threshold; and

if the combination does not exceed the probability threshold, flagging the metadata associated with the media data file for reevaluation.

8. The computer accessible medium of claim 7, the method further comprising:

determining an attribute dependency for the first attribute;

determining whether the metadata includes a third attribute that satisfies the dependency; and

if the dependency is not satisfied, flagging the metadata associated with the media data file for reevaluation.

9. The computer accessible medium of claim 7, the method further comprising:

determining whether a total number of attributes defined in the metadata associated with the media data file exceeds a predetermined value; and

if the total number of attributes does not exceed the predetermined value, flagging the metadata associated with the media data file for reevaluation.

10. The computer accessible medium of claim 7, the method further comprising:

searching the metadata associated with the media data file for an expected attribute dimension, the attribute dimension associated with a subset of attributes; and

if the expected attribute dimension is missing, flagging the metadata associated with the media data file for reevaluation.

11. The computer accessible medium of claim 10, the method further comprising:

determining whether the expected attribute dimension includes a threshold number of defined attributes in the subset; and

if the expected attribute dimension does not include the threshold number of defined attributes in the subset, flagging the metadata associated with the media data file for reevaluation.

12. The computer accessible medium of claim 7, the method further comprising:

searching the metadata associated with the media data file for an expected attribute; and

if the expected attribute is missing, flagging the metadata associated with the media data file for reevaluation.

13. A system for controlling the quality of metadata associated with a media data file, the system comprising:

a media content database comprising a plurality of media data files and respective metadata; and

a quality analysis component configured to;

select a set of metadata associated with a particular media data file from the media content database;

select a first attribute from the set of metadata;

determine an attribute dependency for the first attribute;

determine whether the set of metadata includes a second attribute that satisfies the dependency; and

if the dependency is not satisfied, flag the metadata associated with the media data file for reevaluation.

14. The system of claim 13, wherein the quality analysis component is further configured to:

select a third attribute from the set of metadata;

determine whether the combination of the first attribute with the third attribute exceeds a probability threshold; and

if the combination does not exceed the probability threshold, flag the set of metadata for reevaluation.

15. The system of claim 13, wherein the quality analysis component is further configured to:

determine whether a total number of attributes defined in the set of metadata exceeds a predetermined value; and

if the total number of attributes does not exceed the predetermined value, flag the set of metadata for reevaluation.

16. The system of claim 13, wherein the quality analysis component is further configured to:

search the set of metadata for an expected attribute dimension, the attribute dimension associated with a subset of attributes; and

if the expected attribute dimension is missing, flag the set of metadata for reevaluation.

17. The system of claim 16, wherein the quality analysis component is further configured to:

determine whether the expected attribute dimension includes a threshold number of defined attributes in the subset; and

if the expected attribute dimension does not include the threshold number of defined attributes in the subset, flag the set of metadata for reevaluation.

18. The system of claim 13, wherein the quality analysis component is further configured to:

search the set of metadata for an expected attribute; and

if the expected attribute is missing, flag the set of metadata for reevaluation.

19. The system of claim 13, further comprising a categorization component configured to:

receive an attribute selection from a user, the received attribute selection corresponding to a new media data file;

based on the received attribute selection, determine a group of dependent attributes corresponding to the new media data file that exceed a predetermined probability threshold for the selected attribute;

display the group of dependent attributes to the user; and

allow the user to select from the group of dependent attributes.

20. The system of claim 19, wherein the categorization component is further configured to:

determine whether the received attribute selection is consistent with previously defined attributes corresponding to the new media data file; and

if the received attribute selection is inconsistent with the previously defined attributes, reject the selected attribute.

21. The system of claim 19, wherein the categorization component is further configured to:

determine that the received attribute selection corresponds to an attribute dimension that allows a single attribute to be selected from a subset of attributes;

determine that another attribute from the subset has already been selected; and

automatically assign either the received attribute selection or the other attribute from the subset based on the attribute with the highest probability of occurring within the set of metadata.

22. A method for controlling the quality of metadata associated with a media data file, the method comprising:

receiving an attribute selection from a user, the received attribute selection corresponding to a new media data file;

based on the received attribute selection, determining a group of dependent attributes corresponding to the new media data file that exceed a predetermined probability threshold for the selected attribute;

displaying the group of dependent attributes to the user; and

allowing the user to select from the group of dependent attributes.

23. The method of claim 22, further comprising:

determining whether the received attribute selection is consistent with previously defined attributes corresponding to the new media data file; and

if the received attribute selection is inconsistent with the previously defined attributes, rejecting the selected attribute.

24. The system of claim 23, further comprising:

determining that the received attribute selection corresponds to an attribute dimension that allows a single attribute to be selected from a subset of attributes;

determining that another attribute from the subset has already been selected; and

automatically assigning either the received attribute selection or the other attribute from the subset to the attribute dimension based on the attribute with the highest probability of occurring within the set of metadata.

25. A system comprising:

means for scanning a media content database for inconsistent attribute combinations in metadata corresponding to respective media data files;

means for monitoring inconsistent attribute selections corresponding to a new media data file, the attribute selection received from a user; and

means for flagging any detected attribute combinations and attribute selections for further evaluation.