System to generate related search queries

• US 8,200,687 B2
• Filed: 12/30/2005
• Issued: 06/12/2012
• Est. Priority Date: 06/20/2005
• Status: Active Grant

• Current Assignee: PayPal, Inc. (PayPal Holdings, Inc.)
• Original Assignee: eBay Inc.
• Inventors: Gupta, Raghav
• Primary Examiner(s): Vital, Pierre
• Assistant Examiner(s): Rostami, Mohammad S
• Application Number:
  US11/323,486
• Publication Number:
  US 20060288000A1
• Time in Patent Office: 2,356 Days
• Field of Search: None
• US Class Current: 707/767
• CPC Class Codes: G06F 16/2425 Iterative querying; Query formulation based on the results of a preceding query
• CPC Class Codes: G06F 16/90324 using system suggestions
• CPC Class Codes: G06F 16/9535 Search customisation based on user profiles and personalisation

Abstract

System and methods are described to provide new recommendations to a search interface to assist users in navigating toward new searches that are likely to generate results aligned with the user'"'"'s searching intentions. An algorithm analyzes previous search sessions to determine additional locations for the user to go. In an example of a commerce system, new information placement may be added to the top of search and listings pages to show links to new searches that can be run by the user. On a periodic basis, the search recommendations may be updated, for example based on the most current changes in user behavior.

Claims

1. A system comprising:

memory;

and a processor operatively coupled to the memory, wherein the processor is operable to run;

a search engine to execute a database search based upon a current user determined search query A; and

recommendation engine to analyze prior searches performed for other users to;

determine that one or more prior users that performed search query A, then performed search query B, and then performed predetermined post-search activity with respect to a result set returned by the search query B, wherein the search query'"'"' B comprises a search query that does not include an), terminology in common with the search query A; and

recommend the search query B to the current user,

wherein the recommendation engine is configured to determine whether the search query B includes a search operator and to refrain From recommending the search query B to the current user upon determination that the search query B includes the search operator, wherein the search operator comprises parentheses or Boolean operators; and

wherein the recommendation engine is configured to count a first number of characters of keywords in the search query B and a second number of characters of a category constraint for the search query B and to refrain from recommending the search query B to the current user based in determination that a sum of the first and second numbers exceeds a specified number.

2. The system of claim 1 wherein the recommendation engine analyzes a search query history of the other users and determines a success level in a predetermined activity of the other users.

3. The system of claim 2 wherein the predetermined activity is a commerce transaction.

4. The system of claim 3 wherein the commerce transaction is placing a bid in an auction.

5. The system of claim 1 wherein the recommendation engine determines a rank for one or more recommended searches based on post-search user or system behavior with respect to the database and the prior searches performed for the other users.

6. The system of claim 5 wherein the user behavior indicates that subsequent to each of the one or more recommended searches a threshold number of users purchased one or more products/services identified by the associated recommended search.

7. The system of claim 1 wherein the recommendation engine does not provide one or more recommended searches for the current user if the current user determined search query A includes an identified grammatical error.

8. The system of claim 1 wherein the current user determined search query A includes a product title, and the recommendation engine filters one or more recommended searches to avoid displaying recommendations that overlap with the product title.

9. The system of claim 1 wherein the post-search activity comprises purchasing one or more products/services identified by the search query B.

19. The system of claim 1, wherein the recommendation engine is configured to refrain from updating a cookie upon determination that the sum of the first and second numbers exceeds the specified number.

20. The system of claim 1, wherein the recommendation engine is configured to:

determine whether a category constraint is the same for both the search query A and the search query B; and

refrain from recommending the search query B to the current user based on a determination that the category constraint is not the same for both the search query A and the search query B.

10. A computer-implemented method of making search recommendations comprising:

accessing a database;

determining a first number of prior users that performed a database search using search query A;

determining a second number as a subset of the first number of prior users that performed a database search using starch query B immediately subsequent to performing the database search using the search query A, and performed a predetermined post-search activity with respect to a result set returned by the search query B, wherein the starch query B comprises a search query that does not include any terminology in common with the search query A; and

recommending, upon determination that the second number is larger than zero, the search query B to a current user executing a database search using the search query A, wherein the recommending includes determining whether the search query B includes a search operator and refraining from recommending the search query B to the current user upon determination that the search query B includes the search operator, wherein the search operator comprises parentheses or Boolean operators,

and wherein the recommending includes counting a first number of characters of keywords in the search query B and a second number of characters of a category constraint for the search query B and refraining from recommending the search query B to the current user based on a determination that a sum of the first and second numbers exceeds a specified number.

11. The computer-implemented method of claim 10 wherein the predetermined post-search activity comprises purchasing one or more products/services.

12. A machine-readable medium including a set of instructions that, when executed by a machine, cause the machine perform the method of claim 10.

13. A computer-implemented method of managing a computerized commerce system where a user interface includes page views obtained during a session in response to user specified search queries, the method comprising:

tracking session data to determine sequential page views, wherein page view B is viewed immediately after page view A and wherein post search user activity is done with respect to the page view B, wherein the page view A and the page view B are viewed as a result of search query A and search query B, respectively, wherein the search query B comprises a search query that does not include any terminology in common with the search query A;

editing the tracked session data based upon search queries including a search phrase and a category constraint used to view the page views A and B, wherein the editing comprises eliminating the tracked session data if the category constraint of the search queries is not the same for both the page views A and B, or if the search phrase of the search queries is the same for both for page views A and B, wherein the editing comprises determining whether the search phrase includes a search operator and eliminating the tracked session data upon determination that the search phrase includes the search operator, wherein the search operator comprises parentheses or Boolean operators,

and wherein the editing comprises counting a first number of characters of keywords in the search query B and a second number of characters of a category constraint for the search query B and refraining from recommending the search query B to the current user based on a determination that a sum of the first and second numbers exceeds a specified number; and

maintaining a historical count of a number of times users viewed the page view A followed by the page view B based upon the edited tracked session data for use in recommending the search query B in response to receipt of the search query A from a current user.

14. The computer-implemented method of claim 13 wherein editing the tracked session data based upon a search phrase and a category constraint further comprises eliminating the tracked session data if the category constraint is on a prohibited list.

15. The computer-implemented method of claim 13 wherein tracking session data further comprises tracking predetermined post-search user activity with respect to the page view B.

16. The computer-implemented method of claim 15 wherein the predetermined post-search user activity comprises purchasing one or more products/services.

17. The computer-implemented method of claim 13 further comprises recommending a search query to view the page view B in response to a user specified search query resulting in viewing the page view A.

21. The computer-implemented method of claim 13, wherein the eliminating the tracked session data comprises eliminating sequential page views by the page view A followed by the page view B.

18. A computer-implemented method of managing a computerized commerce system having a user interface including page views obtained during a session in response to specified search queries, the method comprising:

associating page views by prior users, wherein a first page view is associated with an immediately subsequent second page view based upon viewing frequency and post search user activity with respect to the second page view, wherein the first page view and the second page view are viewed as a result of search query A and search query B, respectively, wherein the search query B comprises a search query that does not include any terminology in common with the search query A;

receiving from a current user a user specified search query producing a search result including the first page view; and

recommending a system specified search query producing a search result including the second page view in response to receipt of the user specified search query, wherein the recommending includes determining whether the search query B includes a search operator and refraining from recommending the search query B to the current user upon determination that the search query B includes the search operator, wherein the search operator comprises parentheses or Boolean operators,

and wherein the recommending includes counting a first number of characters of keywords in the search query B and a second number of characters of a category constraint for the search query B and refraining from recommending the search query B to the current user based on a determination that a sum of the first and second numbers exceeds a specified number.