System and method of collaborative filtering based on attribute profiling

• US 8,001,008 B2
• Filed: 12/11/2006
• Issued: 08/16/2011
• Est. Priority Date: 10/24/2006
• Status: Active Grant

• Current Assignee: Nytell Software LLC (Intellectual Ventures LLC)
• Original Assignee: Garett Engle
• Inventors: Engle, Garett
• Primary Examiner(s): ZURITA, JAMES H
• Application Number:
  US11/636,928
• Publication Number:
  US 20080097867A1
• Time in Patent Office: 1,709 Days
• Field of Search: 705/26, 705/37
• US Class Current: 705/26.62
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising
• CPC Class Codes: G06Q 30/0625 Directed, with specific intent or strategy
• CPC Class Codes: G06Q 30/0631 Item recommendations
• CPC Class Codes: G06Q 30/0641 Shopping interfaces
• CPC Class Codes: G06Q 40/04 Trading; Exchange, e.g. stocks, commodities, derivatives or currency exchange

Abstract

A system and/or a method of collaborative filtering based on attribute profiling is disclosed. In one embodiment, a method includes generating a product profile of a content-based product through applying any number of value data assigned by a user to a set of attributes embodying possible reasons as to why the user reacts to the content-based product, generating a user profile of a user through applying a group of value data assigned by the user to a content-based product sharing the set of attributes and recommending a different content-based product matches with the user profile beyond a threshold value. The method may include recommending the different content-based product to a different user when a different user profile of the different matches with the user profile of the user who has subscribed to the different content-based product.

Claims

1. A method, comprising:

collecting a plurality of value data for storage in memory, the plurality of value data assigned by at least one user to a set of attributes embodying possible reasons as to why the at least one user reacts to the content-based product, wherein the value data is selected from the group comprising an indication of liking the content-based product, an indication of no opinion on the content-based product, and an indication of disliking the content-based product; and

executing instructions stored in memory, wherein execution of the instructions by a processor;

generates a product profile of a content-based product through applying the collected plurality of value data;

generates a user profile of the at least one user through applying a group of value data assigned by the at least one user to at least one content-based product sharing the set of attributes; and

generates a recommendation for a different content-based product to the user when a different product profile of the different content-based product corresponds with the user profile beyond a threshold value.

2. The method of claim 1, further comprising generating a score for the content-based product, wherein the indication of liking the content-based product results in adding to the score, the indication of no opinion on the content-based product results in the score remaining the same, and the indication of disliking the content-based product results in subtracting from the score.

3. The method of claim 2, further comprising recommending the different content-based product to a different user when a different user profile of the different user corresponds with the user profile of the user who has subscribed to the different content-based product.

4. The method of claim 2, further comprising recommending a particular content-based product to the user who has subscribed to the different content-based product when a particular product profile of the particular content-based product corresponds with the different product profile of the different content-based product.

5. The method of claim 2, further comprising simultaneously rendering to the at least one user at least one possible genre of the content-based product and at least one set of attributes with each set of attributes corresponding to each genre of the at least one possible genre of the content-based product.

6. The method of claim 5, further comprising applying a bio data of the user to perform the recommending the different content-based product so as to increase an accuracy of the recommending.

7. The method of claim 6, further comprising determining the set of attributes to display to the at least one user based on a plurality of tag data assigned by a group of opinion leaders evaluating the content-based product prior to the display.

8. The method of claim 7, further comprising generating a discrete user profile associated with each genre of the at least one content-based product, wherein the each genre is determined by at least one of a distinctive style, a form, and a content of the at least one content-based product.

9. The method of claim 8, further comprising:

searching for a set of text data reviewing the content-based product through a plurality of selected websites to collect a plurality of meta-data describing the content-based product; and

evaluating the plurality of meta-data to determine a genre of the content based product.

10. The method of claim 9, further comprising selecting a collection of attributes from the plurality of meta-data based on an algorithm which considers at least a frequency of occurrence and a semantics of each of the plurality of meta-data.

11. The method of claim 10, further comprising rendering to the at least one user the collection of attributes associated with the genre of the content-based product when the at least one user selects the genre.

12. The method of claim 1, further comprising communicating an invitation data to an unregistered user acquainted by a registered user.

13. A machine-readable storage medium embodying a set of instructions that, when executed by a machine, causes the machine to perform a method comprising:

collecting a plurality of value data for storage in memory, the plurality of value data assigned by at least one user to a set of attributes embodying possible reasons as to why the at least one user reacts to the content-based product, wherein the value data is selected from the group comprising an indication of liking the content-based product, an indication of no opinion on the content-based product, and an indication of disliking the content-based product;

generating a product profile of a content-based product through applying the collected plurality of value data;

generating a user profile of the at least one user through applying a group of value data assigned by the at least one user to at least one content-based product sharing the set of attributes; and

generating a recommendation for a different content-based product to the user when a different product profile of the different content-based product corresponds with the user profile beyond a threshold value.

14. A system, comprising:

a content provide module stored in memory and executable by a processor to collect a plurality of value data assigned by at least one user to a set of attributes embodying possible reasons as to why the at least one user react to at least one content-based product, wherein the value data is selected from the group comprising an indication of liking the content-based product, an indication of no opinion on the content-based product, and an indication of disliking the content-based product;

an attribute profile module stored in memory and executable by the processor to generate;

a product profile of a content-based product through applying the collected plurality of value data, and

a user profile of the at least one user through applying a group of value data assigned by the at least one user to at least one content-based product sharing the set of attributes; and

a recommendation module stored in memory and executable by the processor to recommend the at least one content-based product to a particular user when a product profile of the at least one content-based product corresponds with a user profile of the particular user beyond a threshold value.

15. The method system of claim 14, further comprising an attribute select module stored in memory and executable by the processor to fetch and displaying the set of attributes corresponding to the at least one content-based product.

16. The system of claim 15, further comprising a match module stored in memory and executable by the processor to connect the particular user to the at least one content-based product when the user profile of the particular corresponds with the product profile of the at least one-content based product beyond a threshold value.