Method and System for High Performance Model-Based Personalization

• US 20100332408A1
• Filed: 08/30/2010
• Published: 12/30/2010
• Est. Priority Date: 06/23/2000
• Status: Active Grant

• Current Assignee: Thalveg Data Flow LLC (Intellectual Ventures LLC)
• Original Assignee: Thalveg Data Flow LLC (Intellectual Ventures LLC)
• Inventors: EKHAUS, Michael A., Driskill, Robert, Mulier, Filip
• Granted Patent: US 8,155,992 B2
• Time in Patent Office: Days
• US Class Current: 705/347
• CPC Class Codes: G06F 16/2462 Approximate or statistical queries
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising
• CPC Class Codes: G06Q 30/0201 Market modelling; Market analysis; Collecting market data
• CPC Class Codes: G06Q 30/0282 Rating or review of business operators or products
• CPC Class Codes: G06Q 30/0631 Item recommendations

Abstract

The present invention relates to a method and system for generating client preference recommendations in a high performance computing regime. Accordingly, one embodiment of the present invention comprises: providing a sparse ratings matrix, forming a plurality of data structures representing the sparse ratings matrix, forming a runtime recommendation model from the plurality of data structures, determining a recommendation from the runtime recommendation model in response to a request from a user, and providing the recommendation to the user.

Claims

1. A method for preparing a recommendation comprising:

providing, in memory, a sparse unary ratings matrix from users'"'"' selected preferences, wherein each user'"'"'s selected preferences are represented as unary data entries in the sparse unary ratings matrix, wherein each unary data entry has a value of either zero or one;

banding by a data processing device the sparse unary ratings matrix, wherein bands of the sparse unary ratings matrix partition ratings by client;

distributing the bands to a plurality of computing nodes, wherein the plurality of computing nodes generate from the bands respective output that together form a matrix of co-rates, wherein the matrix of co-rates includes either a pre-multiplication of the sparse unary ratings matrix by a transpose of the sparse unary ratings matrix or a post-multiplication of the sparse unary ratings matrix by the transpose of the sparse unary ratings matrix;

forming in the data processing device a runtime recommendation model from the output of the plurality of computing nodes;

determining in the data processing device a recommendation from the runtime recommendation model in response to a request from a user; and

providing the recommendation to the user.

2. The method of claim 1, further comprising:

calculating a unary multiplicity voting recommendation from the runtime recommendation model.

3. The method claim 1, further comprising:

calculating a non-unary multiplicity voting recommendation from the runtime recommendation model.

4. The method of claim 2, wherein said calculating a unary multiplicity voting recommendation comprises calculating an anonymous recommendation.

5. The method of claim 2, wherein said calculating a unary multiplicity voting recommendation comprises calculating a personalized recommendation.

6. The method of claim 3, wherein said calculating a non-unary multiplicity voting recommendation comprises calculating an anonymous recommendation.

7. The method of claim 3, wherein said calculating a non-unary multiplicity voting recommendation comprises calculating a personalized recommendation.

8. A method for preparing a recommendation comprising:

providing, in memory, a sparse unary ratings matrix from users'"'"' selected preferences, wherein each user'"'"'s selected preferences are represented as unary data entries in the sparse unary ratings matrix, wherein each unary data entry has a value of either zero or one;

striping by a data processing device the sparse unary ratings matrix, wherein stripes of the sparse unary ratings matrix partition ratings by item;

distributing the stripes to a plurality of computing nodes, wherein the plurality of computing nodes generate from the stripes respective output that together form a matrix of co-rates, wherein the matrix of co-rates includes either a pre-multiplication of the sparse unary ratings matrix by a transpose of the sparse unary ratings matrix or a post-multiplication of the sparse unary ratings matrix by the transpose of the sparse unary ratings matrix;

forming in the data processing device a runtime recommendation model from the output of the plurality of computing nodes;

determining in the data processing device a recommendation from the runtime recommendation model in response to a request from a user; and

providing the recommendation to the user.

9. The method of claim 8, further comprising:

calculating a unary multiplicity voting recommendation from the runtime recommendation model.

10. The method claim 8, further comprising:

calculating a non-unary multiplicity voting recommendation from the runtime recommendation model.

11. The method of claim 9, wherein said calculating a unary multiplicity voting recommendation comprises calculating an anonymous recommendation.

12. The method of claim 9, wherein said calculating a unary multiplicity voting recommendation comprises calculating a personalized recommendation.

13. The method of claim 10, wherein said calculating a non-unary multiplicity voting recommendation comprises calculating an anonymous recommendation.

14. The method of claim 10, wherein said calculating a non-unary multiplicity voting recommendation comprises calculating a personalized recommendation.

15. A method of preparing a recommendation comprising:

providing, in memory, a first ratings matrix associated with a first category of data;

providing, in memory, a second ratings matrix associated with a second category of data, wherein the first category of data and the second category of data each represent mappings between differing sources of ratings data;

forming in a data processing device a runtime recommendation model from a cross-set of co-occurrences of the first ratings matrix and the second ratings matrix;

calculating in the data processing device a recommendation from the runtime recommendation model in response to a request from a user; and

providing the recommendation to the user.

16. The method of claim 15, further comprising:

calculating a unary multiplicity voting recommendation from the runtime recommendation model.

17. The method claim 15, further comprising:

calculating a non-unary multiplicity voting recommendation from the runtime recommendation model.

18. The method of claim 16, wherein said calculating a unary multiplicity voting recommendation comprises calculating an anonymous recommendation.

19. The method of claim 16, wherein said calculating a unary multiplicity voting recommendation comprises calculating a personalized recommendation.

20. The method of claim 17, wherein said calculating a non-unary multiplicity voting recommendation comprises calculating an anonymous recommendation.

21. The method of claim 17, wherein said calculating a non-unary multiplicity voting recommendation comprises calculating a personalized recommendation.