Systems and methods of determining genre information

• US 8,204,883 B1
• Filed: 04/17/2008
• Issued: 06/19/2012
• Est. Priority Date: 04/17/2008
• Status: Active Grant

• Current Assignee: Amazon Technologies, Inc. (Amazon.com, Inc.)
• Original Assignee: Amazon Technologies, Inc. (Amazon.com, Inc.)
• Inventors: Tams, Jonathan G. G., Smith, Graeme N., Round, Matthew J., Seymour, Dominic E. H. K.
• Primary Examiner(s): Vital, Pierre
• Assistant Examiner(s): Lee, Rachel J
• Application Number:
  US12/105,163
• Time in Patent Office: 1,524 Days
• Field of Search: None
• US Class Current: 707/736
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising

Abstract

Systems and methods of determining genre information are provided. A particular method includes receiving a first data set including a plurality of genre descriptors and associated strength values related to particular content. Each strength value indicates a degree to which the associated genre descriptor describes the particular content. The method also includes storing a data record associating the first data set with the particular content.

Claims

1. A method, comprising:

acquiring data sets that include a plurality of strength values and a plurality of classification descriptors for an entity that produced a particular content, wherein each strength value indicates a degree to which an associated classification descriptor describes the entity;

sorting the classification descriptors into a plurality of categories, each category storing a unique combination of a classification descriptor and an associated strength value;

determining a score for each category of the plurality of categories, each score being calculated based at least on a number of times that a corresponding category appears in the data sets; and

for each set of categories that have a unique strength value;

obtaining a cut-off threshold by dividing a sum of scores of corresponding categories by a number of the corresponding categories; and

selecting each of one or more classification descriptors having an associated score that is above the cut-off threshold as a designated classification descriptor for the entity.

2. The method of claim 1, wherein the determining includes determining a score for a category by multiplying an strength value of the category by the number of times that the category appears in the data sets.

3. The method of claim 1, further comprising:

querying at least one user to identify a degree to which a plurality of genre descriptors describe the particular content;

receiving additional data sets including the plurality of genre descriptors and associated additional strength values related to the particular content, wherein each additional strength value indicates the degree to which an associated genre descriptor describes the particular content; and

storing a data record associating the data set with the particular content.

4. The method of claim 3, further comprising determining a genre vector based at least partially on the plurality of genre descriptors, wherein the genre vector is descriptive of a coordinate location of the particular content in a genre space that includes an axis associated with each of the plurality of genre descriptors, wherein a coordinate along each axis is determined based on a particular strength value of the additional data sets.

5. The method of claim 4, further comprising determining an entity genre vector of the entity based at least partially on the genre vector.

6. The method of claim 5, wherein the entity genre vector is further determined by averaging the genre vector with one or more additional genre vectors of other content associated with the entity.

7. The method of claim 1, wherein the entity comprises a collection of content including the particular content and the other content.

8. The method of claim 1, wherein the entity comprises at least one of an artist, a group of artists, an author, a production company or a distribution company.

9. The method of claim 1, wherein the particular content comprises at least one of a performance, a literary work, and an artistic work.

10. The method of claim 5, further comprising quantizing the entity genre vector to determine a representative entity genre vector, wherein the acquiring includes determining based on the representative entity genre vector the plurality of genre descriptors for the entity.

11. A system, comprising:

one or more processors;

a memory that includes a plurality of components that comprises;

a genre database to store genre information for a plurality of content items, wherein the genre information includes a genre vector for each content item, wherein the genre vector indicates a degree to which each of a plurality of genre descriptors that are received from one or more users describe each content item; and

a genre mapping module to;

analyze a plurality of genre vectors to determine a plurality of classification data sets for a content item, wherein each classification data set includes a strength descriptor and a genre descriptor, wherein the strength descriptor describes how relevant the genre descriptor is to the content item;

ascertain strength values of the strength descriptors associated with each genre descriptor of the plurality of classification data sets to determine a most frequently occurring strength value;

designate the most frequently occurring strength value as a representative strength value for the each genre descriptor;

modify each strength descriptor for the each genre descriptor that has a higher strength value than the representative strength value to assume the representative strength value; and

output one or more of the plurality of classification data sets that include at least one strength descriptor with a modified strength value.

12. The system of claim 11, wherein the genre mapping module is to further determine a coordinate location of a particular content item in a multi-dimensional space based on a corresponding genre vector, and modify the coordinate location based on weights assigned to the plurality of genre descriptors, wherein the multidimensional space is defined by a plurality of axes and each axis of the plurality of axes is associated with a particular genre descriptor.

13. The system of claim 12, wherein the plurality of components further comprises a location comparator to compare a first coordinate location of a first content item of the plurality of content items with a second coordinate location of a second content item of the plurality of content items.

14. The system of claim 13, wherein the location comparator is to compare the coordinate location of the first content item and the coordinate location of the second content item at least by:

quantizing a first genre vector and identifying a first representative genre vector based on the quantized first genre vector, wherein the first genre vector is associated with the first coordinate location;

quantizing a second genre vector and identifying a second representative genre vector based on the quantized second genre vector, wherein the second genre vector is associated with the second coordinate location; and

estimating a distance between the first representative genre vector and the second representative genre vector.

15. The system of claim 14, wherein the location comparator is to further determine relative distances from the first coordinate location to coordinate locations associated with the plurality of content items, and to identify content items related to the first content item based on the relative distances.

16. The system of claim 15, wherein the relative distances are used to simplify calculation of distances between the first coordinate location and the coordinate locations associated with the plurality of content items.

17. The system of claim 12, further comprising an exemplar selection module to select one content item of a plurality of content items having the same degree of association to a particular genre descriptor as a representative content item for the particular genre descriptor when the one content item is more popular than the other content item of the plurality of content items.

18. The system of claim 17, further comprising a display interface to output display data to generate a display screen including a plurality of representative content items, wherein the plurality of representative content items are ordered in the display screen based at least partially on relative popularity of the plurality of content items.

19. The system of claim 17, wherein the exemplar selection module identifies at least one exemplar content item for a particular genre descriptor by identifying at least one content item with a genre vector nearer to a particular axis of the plurality of axes than other genre vectors, wherein the particular axis is associated with the particular genre descriptor.

20. The system of claim 12, wherein the plurality of components further comprises an aggregator module to aggregate data received from a plurality of sources related to a particular content item and to determine the genre vector for the particular content item based on the aggregated data.

21. The system of claim 12, wherein the plurality of components further comprises an aggregator module to aggregate data related to each of the plurality of content items, wherein the each of the plurality of content items is associated with a particular entity, and to determine an entity genre vector for the particular entity based on the aggregated data.

22. The system of claim 21, wherein the genre mapping module is further to determine a coordinate location of a particular entity in a multidimensional space defined by a plurality of axes based on an entity genre vector associated with the particular entity, wherein each axis of the plurality of axes is associated with a particular genre descriptor.

23. The system of claim 22, further comprising a location comparator module to compare a first coordinate location of a first entity with a second coordinate location of a second entity.

24. The system of claim 23, wherein the location comparator module is to further weight the first coordinate location of the first entity and the second coordinate location of the second entity using a frequency vector before comparing the first coordinate location of the first entity and the second coordinate location of the second entity, wherein the frequency vector includes information about a frequency with which each genre descriptor of the plurality of genre descriptors is associated with entities in the genre database.

25. The system of claim 21, wherein the plurality of components further comprises an exemplar selection module to select at least one representative entity for each of the plurality of genre descriptors, wherein the at least one representative entity for a particular genre descriptor has a large degree of association to the particular genre descriptor.

26. A method comprising:

accessing a plurality of classification vectors, each classification vector including strength data indicating a degree of relevance of a particular classification descriptor to an entity;

analyzing the plurality of classification vectors to determine a plurality of classification data sets for the entity, wherein each classification data set includes a strength descriptor and a classification descriptor, wherein the strength descriptor describes how relevant the classification descriptor is to the entity;

ascertaining strength values of the strength descriptors associated with each classification descriptor of the plurality of classification data sets to determine a most frequently occurring strength value;

designating the most frequently occurring strength value as a representative strength value for the each classification descriptor;

modifying each strength descriptor for the each classification descriptor that has a higher strength value than the representative strength value to assume the representative strength value; and

outputting one or more of the plurality of classification data sets that include at least one strength descriptor with a modified strength value.

27. The method of claim 26, wherein the classification descriptor comprises a description of a particular genre.

28. The method of claim 26, wherein each strength value indicate a degree of relevance of a corresponding classification descriptor to the entity.

29. The method of claim 26, wherein the modifying increases a number of classification descriptors with the representative strength value.

30. The method of claim 26, wherein the plurality of classification vectors are provided by users, wherein each of the users is associated with a user reputation score, and wherein the representative strength descriptor is modified based at least partially on the user reputation scores associated with the users.

31. The method of claim 26, wherein each strength descriptor comprises an adjective indicating a relative relevance of the classification descriptor to the entity.

32. The method of claim 31, wherein the strength descriptor comprises at least one of hints of, mainly, and quite.

33. A method, comprising:

receiving a data set from a user, wherein the data set is associated with a particular entity and the data set includes a classification descriptor of a plurality of classification descriptors and an associated strength descriptor of a plurality of strength descriptors;

storing the data set with a number of additional data sets associated with the particular entity to produce an aggregated data set, each additional data set including a classification descriptor of the plurality of classification descriptors and an associated strength descriptor of the plurality of strength descriptors;

sorting the aggregated data set into a number of categories, wherein each category is associated with a particular classification descriptor and a particular strength descriptor;

determining a score for each category of the number of categories, wherein the score is based on a number of data sets associated with a respective category and a strength value assigned to the particular strength descriptor of the respective category;

determining a number of candidate classification descriptors assigned to a specific strength descriptor by analyzing the classification descriptors in the aggregated data set;

rejecting a candidate classification descriptor that is last to be stored in the aggregated data set from consideration when the number of candidate classification descriptors previously assigned to the specific strength descriptor exceeds a maximum number of classification descriptors for the specific strength descriptor; and

designating each of the candidate classification descriptors as a designated category classification for the particular entity when the number of candidate classification descriptors assigned to the specific strength descriptor does not exceed the maximum number of classification descriptors for the specific strength descriptor.

34. The method of claim 33, wherein the score associated with a particular category is based on one or more user reputation values, and wherein each of the one or more user reputation values is associated with a respective data set associated with the particular category.

35. The method of claim 33, wherein the rejecting the candidate classification descriptor includes rejecting the classification descriptor that is included in the data set.

36. The method of claim 33, wherein the number of candidate classification descriptors is determined by comparing the scores associated with the categories related to each particular classification descriptor with the scores associated with the categories related to one or more additional classification descriptors.

37. The method of claim 33, further comprising filtering the number of candidate classification descriptors prior to determining the number of candidate classification descriptors assigned to the specific strength descriptor.

38. The method of claim 37, wherein the number of candidate categories is filtered based one or more cut-off thresholds, wherein each cut-off threshold related to a particular entity is associated with the strength descriptor of the particular entity.

39. The method of claim 33, further comprising providing a user interface display indicating the particular entity and one or more designated categories associated with the particular entity.

40. The method of claim 38, wherein a cut-off threshold is determined by dividing a sum of strength descriptors for the number of candidate classification descriptors assigned to the specific strength descriptor by the number of candidate classification descriptors.