User interactive precision targeting principle

• US 7,882,111 B2
• Filed: 06/01/2007
• Issued: 02/01/2011
• Est. Priority Date: 06/01/2007
• Status: Active Grant

• Current Assignee: R2 Solutions LLC (Acacia Research Corporation)
• Original Assignee: Yahoo! Inc. (Apollo Global Management, Inc.)
• Inventors: John, George H., Rochnik, Nikolai, Chien, Shu-Yao, Seth, Amitabh
• Primary Examiner(s): Mahmoudi; Tony
• Assistant Examiner(s): Uddin; Md. I
• Application Number:
  US11/757,296
• Publication Number:
  US 20080301118A1
• Time in Patent Office: 1,341 Days
• Field of Search: 707/5, 707/2, 707/4, 707/728, 707/748, 707/749, 707/750, 707/751, 707/758, 705/27
• US Class Current: 707/749
• CPC Class Codes: G06F 16/954 Navigation, e.g. using categorised browsing

Abstract

A method of determining content relevance for a user sets a user preference, which is related to a first area of content. The method calculates a set of scores, by using a combination, of the user preference, affinity data, and a parametric weight. The method organizes the content by using the set of scores, such that the organization of the content has a desirable relationship to the user, and recommends the selected content. Preferably, the method precomputes the affinity data and/or the parametric weight to generate and store the precompiled data for later retrieval. The affinity data describes a relationship between a first item of content and a second item of content, and the parametric weight describes an attribute of the second item. Additional embodiments include a system implementation and computer readable medium.

Claims

1. A computer-implemented method for determining relevance for a user between a predictor item of content and a plurality of target items of content, said method comprising:

storing, in data storage, at least one weight parameter for said target items that measures similarity of content between a target item and said predictor item;

storing an affinity score for each of said target items relative to said predictor item, that measures a correlation between said predictor item and said corresponding target item based on interests of a plurality of users;

setting at least one user preference parameter that indicates similarity, including more similar and less similar, desired by said user for content relative to said predictor item;

processing, in a computer, to determine relevance of said target items of content for said user by generating a final score by;

multiplying said weight parameters on similarity for said target items with said user preference parameter for more similar or less similar to generate a user modified similarity score;

adding said affinity score for each of said target items to said modified similarity score for each of said target items to generate said final score for each of said target items; and

determining relevance of said target items of content for said user based on said final score for said target items.

2. The method as set forth in claim 1, further comprising:

storing an affinity score for said target items of content that identifies a relationship for user affinity of content between a target item and a predictor item; and

determining relevance of said target items of content for said user based on said affinity score.

3. The method as set forth in claim 2, further comprising updating one of said user preference, said weight parameter or said affinity score.

4. The method as set forth in claim 1, further comprising:

storing, in data storage, a plurality of categories with said target items classified in said categories, so as to depict distance relationships among said target items of content; and

determining similarity between said predictor item and said target items based on said distance relationships.

5. The method as set forth in claim 1, further comprising:

displaying, on a user computer, an interactive user interface that permits a user to set said user preference.

6. The method as set forth in claim 1, wherein setting said user preference comprises:

setting said user preference based on data obtained from a set of demographically similar users.

7. The method as set forth in claim 1, further comprising:

selecting one of said target items of content to deliver to said user.

8. The method as set forth in claim 1, wherein:

setting at least one user preference that indicates similarity comprises setting a range of interest for similarity to indicate user preference.

9. A system for determining relevance for a user between a predictor item of content and a plurality of target items of content, said system comprising:

data storage for storing at least one weight parameter for said target items that measures similarity of content between a target item and said predictor item, and storing an affinity score for each of said target items relative to said predictor item, that measures a correlation between said predictor item and said corresponding target item based on interests of a plurality of users;

at least one computer, coupled to said data storage, for setting at least one user preference parameter that indicates similarity, including more similar and less similar, desired by said user for content relative to said predictor item, for multiplying said weight parameters on similarity for said target items with said user preference parameter for more similar or less similar to generate a user modified similarity score, adding said affinity score for each of said target items to said modified similarity score for each of said target items to generate said final score for each of said target items, and for determining relevance of said target items of content for said user based on said final score for said target items.

10. The system as set forth in claim 9, wherein:

said data storage further for storing an affinity score for said target items of content that identifies a relationship for user affinity of content between a target item and a predictor item; and

said computer further for determining relevance of said target items of content for said user based on said affinity score.

11. The system as set forth in claim 10, wherein said computer further for updating one of said user preference, said weight parameter or said affinity score.

12. The system as set forth in claim 9, wherein:

said data storage further for storing, in data storage, a plurality of categories with said target items classified in said categories, so as to depict distance relationships among said target items of content; and

said computer further for determining similarity between said predictor item and said target items of based on said categories.

13. The system as set forth in claim 9, further comprising:

a user computer for displaying an interactive user interface that permits a user to set said user preference.

14. The system as set forth in claim 9, wherein:

said computer further for setting said user preference based on data obtained from a set of demographically similar users.

15. The system as set forth in claim 9, wherein:

said computer further for selecting one of said target items of content to deliver to said user.

16. The system as set forth in claim 9, wherein:

said computer further for setting a range of interest for similarity to indicate user preference.

17. A non-transitory computer readable medium storing a program for determining relevance for a user between a predictor item of content and a plurality of target items of content, comprising:

storing, in data storage, at least one weight parameter for said target items of content that measures similarity of content between a target item and said predictor item;

storing an affinity score for each of said target items relative to said predictor item, that measures a correlation between said predictor item and said corresponding target item based on interests of a plurality of users;

setting at least one user preference parameter that indicates similarity, including more similar and less similar, desired by said user for content relative to said predictor item;

processing, in a computer, to determine relevance of said target items of content for said user by generating a final score by;

multiplying said weight parameters on similarity for said target items with said user preference parameter for more similar or less similar to generate a user modified similarity score;

adding said affinity score for each of said target items to said modified similarity score for each of said target items to generate said final score for each of said target items; and

determining relevance of said target items of content for said user based on said final score for said target items.

18. The computer readable medium as set forth in claim 17, further comprising:

storing an affinity score for said target items of content that identifies a relationship between a target item and a predictor item; and

determining relevance of said target items of content for said user based on said affinity score.

19. The computer readable medium as set forth in claim 18, further comprising updating one of said user preference, said weight parameter or said affinity score.

20. The computer readable medium as set forth in claim 17, further comprising:

displaying, on a user computer, an interactive user interface that permits a user to set said user preference.

21. The computer readable medium as set forth in claim 17, wherein setting said user preference comprises:

setting said user preference based on data obtained from a set of demographically similar users.

22. The computer readable medium as set forth in claim 17, further comprising:

selecting one of said target items of content to deliver to said user.

23. The computer readable medium as set forth in claim 17, wherein:

setting at least one user preference that indicates popularity comprises setting a range of interest for popularity to indicate user preference.

24. The computer readable medium as set forth in claim 17, further comprising:

storing, in data storage, a plurality of categories with said target items classified in said categories, so as to depict distance relationships among said target items of content; and

determining similarity between said predictor item and said target items based on said distance relationships.