Method and system for generating recommendations

• US 20040176966A1
• Filed: 03/05/2003
• Published: 09/09/2004
• Est. Priority Date: 03/05/2003
• Status: Active Grant

• Current Assignee: Hewlett Packard Enterprise Development LP (Hewlett-Packard Enterprise Company)
• Original Assignee: Hewlett-Packard Development Company, L.P. (HP Inc.)
• Inventors: Chen, Qiming
• Granted Patent: US 8,015,054 B2
• Time in Patent Office: Days
• US Class Current: 705/1
• CPC Class Codes: G06Q 10/087 Inventory or stock management, e.g. order filling, procurement or balancing against orders
• 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 30/0202 Market predictions or forecasting for commercial activities

Abstract

A method and system for generating recommendations are described. The method of generating a recommendation list of products to a customer comprises using OLAP (On-Line Analytical Processing) to analyze raw data in a multilevel and multidimensional manner. Furthermore, the method includes applying a purality of filtering processes to the raw data to generate a plurality of preliminary recommendation lists. Moreover, the method further comprises generating the recommendation list based on the plurality of preliminary recommendation lists.

Claims

1. A method of generating a recommendation list of products to a customer, comprising:

using OLAP (On-Line Analytical Processing) to analyze raw data in a multilevel and multidimensional manner;

applying a purality of filtering processes to said raw data to generate a plurality of preliminary recommendation lists; and

generating said recommendation list based on said plurality of preliminary recommendation lists.

2. The method as recited in claim 1 wherein said plurality of filtering processes includes collaborative filtering (CF).

3. The method as recited in claim 2 further comprising generating a volume cube and a similarity cube.

4. The method as recited in claim 1 wherein said plurality of filtering processes includes information filtering (IF).

5. The method as recited in claim 1 wherein said plurality of filtering processes includes association rule based filtering (RF).

6. The method as recited in claim 5 further comprising generating a volume cube, an association cube, a population cube, a base cube, a confidence cube, and a support cube.

7. The method as recited in claim 1 further comprising:

identifying a neighborhood of said customer using multilevel and multidimensional collaborative filtering;

mining association rules within said neighborhood;

generating a first preliminary recommendation list based on multilevel and multidimensional collaborative filtering; and

adjusting said first preliminary recommendation list according to said neighborhood mined association rules.

8. A computer-readable medium comprising computer-executable instructions stored therein for performing a method of generating a recommendation list of products to a customer, said method comprising:

using OLAP (On-Line Analytical Processing) to analyze raw data in a multilevel and multidimensional manner;

applying a purality of filtering processes to said raw data to generate a plurality of preliminary recommendation lists; and

generating said recommendation list based on said plurality of preliminary recommendation lists.

9. The computer-readable medium as recited in claim 8 wherein said plurality of filtering processes includes collaborative filtering (CF).

10. The computer-readable medium as recited in claim 9 wherein said method further comprises generating a volume cube and a similarity cube.

11. The computer-readable medium as recited in claim 8 wherein said plurality of filtering processes includes information filtering (IF).

12. The computer-readable medium as recited in claim 8 wherein said plurality of filtering processes includes association rule based filtering (RF).

13. The computer-readable medium as recited in claim 12 wherein said method further comprises generating a volume cube, an association cube, a population cube, a base cube, a confidence cube, and a support cube.

14. The computer-readable medium as recited in claim 8 wherein said method further comprises:

identifying a neighborhood of said customer using multilevel and multidimensional collaborative filtering;

mining association rules within said neighborhood;

generating a first preliminary recommendation list based on multilevel and multidimensional collaborative filtering; and

adjusting said first preliminary recommendation list according to said neighborhood mined association rules.

15. A recommendation system comprising:

a data warehouse for storing raw data representing purchases of products by customers;

an OLAP engine for analyzing said raw data in a multilevel and multidimensional manner to generate a recommendation list of products to a customer, wherein said OLAP engine applies a purality of filtering processes to said raw data;

a multidimensional database for storing data cubes processed by the OLAP engine.

16. The recommendation system as recited in claim 15 wherein said plurality of filtering processes includes collaborative filtering (CF).

17. The recommendation system as recited in claim 16 wherein said OLAP engine generates a volume cube and a similarity cube.

18. The recommendation system as recited in claim 15 wherein said plurality of filtering processes includes information filtering (IF).

19. The recommendation system as recited in claim 15 wherein said plurality of filtering processes includes association rule based filtering (RF).

20. The recommendation system as recited in claim 19 wherein said OLAP engine generates a volume cube, an association cube, a population cube, a base cube, a confidence cube, and a support cube.