GENERATING A SEARCH RESULT RANKING FUNCTION

• US 20120059819A1
• Filed: 09/02/2010
• Published: 03/08/2012
• Est. Priority Date: 09/02/2010
• Status: Active Grant

• Current Assignee: PayPal, Inc. (PayPal Holdings, Inc.)
• Original Assignee: eBay Inc.
• Inventors: Severance, Stephen, Wheeler, Paul
• Granted Patent: US 9,405,835 B2
• Time in Patent Office: Days
• US Class Current: 707/723
• CPC Class Codes: G06F 16/24578 using ranking
• CPC Class Codes: G06F 16/248 Presentation of query results
• CPC Class Codes: G06F 16/951 Indexing; Web crawling techniques
• CPC Class Codes: G06F 16/9535 Search customisation based on user profiles and personalisation
• CPC Class Codes: G06F 16/9538 Presentation of query results
• CPC Class Codes: G06Q 30/0625 Directed, with specific intent or strategy

Abstract

A method and a system to provide generate a search result ranking function, processing a search, and presenting search results are described to provide generate a search result ranking function, processing a search, and presenting search results are described. In one embodiment, a genetic algorithm module receives a plurality of factors, a test set of items, and an ordering solution representing the preferred ordering of the test set of items, generates a potential ranking function based on the plurality of factors, and apply the potential ranking function to each item in the test set of items to generate an ordering of items associated with the potential ranking function. The genetic algorithm module also compares the ordering of items with the ordering solution, and identifies, based on the comparison, the potential ranking function as a solution ranking function. A ranking function may assign a ranking score to items in a set of active items, the ranking score based on the solution ranking function.

Claims

10. A system comprising:

a genetic algorithm module to;

receive a plurality of factors, a test set of items, and an ordering solution representing a preferred ordering of the test set of items,

generate a potential ranking function based on the plurality of factors,

apply the potential ranking function to each item in the test set of items to generate an ordering of items associated with the potential ranking function,

compare the ordering of items with the ordering solution, and

identify, based on the result of the comparing, the potential ranking function as a solution ranking function; and

a ranking module to assign a ranking score to items in a set of active items, the ranking scores assigned to the items determined based on the solution ranking function.

1. A method comprising:

receiving a plurality of factors, a test set of items, and an ordering solution representing a preferred ordering of the test set of items;

generating a potential ranking function based on the plurality of factors;

applying, using a processor, the potential ranking function to each item in the test set of items to generate an ordering of items associated with the potential ranking function;

comparing the ordering of items with the ordering solution;

identifying, based on the result of the comparing, the potential ranking function as a solution ranking function; and

assigning a ranking score to items in a set of active items, the ranking scores assigned to the items determined based on the solution ranking function.

2. The method of claim 1, further comprising:

receiving a query;

processing the query to obtain a result set of items that satisfy the query, the result set of items being a subset of the set of active items;

ordering items in the result set of items based on respective ranking scores assigned to the items to generate an ordered result set of items; and

transmitting the ordered result set of items to a computer system associated with a source of the query.

3. The method of claim 2, wherein ranking scores are assigned only to items in the query result list of items.

4. The method of claim 1, further comprising:

assigning, based on a preferred ordering function, a preferred ordering score to each item in the test set of items; and

generating, based on the preferred ordering score of each item in the test set of items, the ordering solution representing the preferred ordering of the test set of items.

5. The method of claim 1, wherein the potential ranking function comprises at least one factor and a weight associated with the at least one factor, the method further comprising:

modifying the potential ranking function by at least one of altering the weight associated with the factor, eliminating the factor, and adding a new factor.

6. The method of claim 1, wherein the potential ranking function comprises at least one factor and a weight associated with the at least one factor, the method further comprising:

modifying the potential ranking function using another ranking function.

7. The method of claim 1, the generating of the potential ranking function further comprising:

selecting a factor from the plurality of factors to be used in the potential ranking function; and

assigning a weight to the factor.

8. The method of claim 7, wherein the factor is randomly selected, and wherein the weight is randomly assigned to the factor.

9. The method of claim 1, wherein the set of active items comprises a plurality of listings for an online trading platform.

11. The system of claim 10, further comprising:

a query processing module to;

receive a query,

process the query to obtain a result set of items that satisfy the query, the result set of items being a subset of the set of active items,

order items in the result set of items based on respective ranking scores assigned to the items to generate an ordered result set of items; and

a query presentation module to transmit the ordered result set of items to a computer system associated with a source of the query.

12. The system of claim 10, further comprising an ordering solution module to:

assign, based on a preferred ordering function, a preferred ordering score to each item in the test set of items, and

generate, based on the preferred ordering score of each item in the test set of items, the ordering solution representing the preferred ordering of the test set of items.

13. The system of claim 10, wherein the genetic algorithm module is further to:

modifying the potential ranking function using another ranking function.

14. The system of claim 10, wherein the genetic algorithm module is further to:

select a factor from the plurality of factors to be used in the potential ranking function; and

assigning a weight to the factor.

12-1. The system of claim 10, wherein the genetic algorithm module is further to:

modify the potential ranking function by at least one of altering a weight associated with a factor, eliminating the factor, and adding a new factor.

15. A machine-readable medium comprising stored instructions, wherein the instructions, when executed, cause a machine to:

receive a plurality of factors, a test set of items, and an ordering solution representing a preferred ordering of the test set of items;

generate a potential ranking function based on the plurality of factors;

apply, using a processor, the potential ranking function to each item in the test set of items to generate an ordering of items associated with the potential ranking function;

compare the ordering of items with the ordering solution;

identify, based on the result of the comparing, the potential ranking function as a solution ranking function; and

assign a ranking score to items in a set of active items, the ranking scores assigned to the items determined based on the solution ranking function.

16. The machine-readable medium of claim 15, further comprising instructions, which when executed, cause a machine to:

receive a query;

process the query to obtain a query result set of items that satisfy the query, the result set of items being a subset of the set of active items;

ordering items in the result set of items based on respective ranking scores assigned to the items to generate an ordered result set of items; and

transmit the ordered result set of items to a computer system associated with a source of the query.

17. The machine-readable medium of claim 16, wherein ranking scores are assigned only to items in the query result list of items.

18. The machine-readable medium of claim 15, further comprising instructions, which when executed, cause a machine to:

assign, based on a preferred ordering function, a preferred ordering score to each item in the test set of items; and

generate, based on the preferred ordering score of each item in the test set of items, the ordering solution representing the preferred ordering of the test set of items.

19. The machine-readable medium of claim 15, further comprising instructions, which when executed, cause a machine to:

modify the potential ranking function by at least one of altering the weight associated with a factor, eliminating the factor, and adding a new factor.

20. The machine-readable medium of claim 15, further comprising instructions, which when executed, cause a machine to:

modify the potential ranking function using another ranking function.