Query suggestion for e-commerce sites

• US 9,323,811 B2
• Filed: 01/27/2015
• Issued: 04/26/2016
• Est. Priority Date: 07/30/2010
• Status: Expired due to Fees

• Current Assignee: eBay Inc.
• Original Assignee: eBay Inc.
• Inventors: Hasan, Mohammad Al, Parikh, Nishith, Singh, Gyanit, Sundaresan, Neelakantan, Johnson, Brian Scott, Khurana, Udayan
• Primary Examiner(s): Hwa, Shyue Jiunn
• Application Number:
  US14/606,971
• Publication Number:
  US 20150142827A1
• Time in Patent Office: 455 Days
• US Class Current: 1/1
• CPC Class Codes: G06F 16/24 Querying
• CPC Class Codes: G06F 16/24578 using ranking
• CPC Class Codes: G06F 16/3322 using system suggestions G06F16/3325 takes precedence
• CPC Class Codes: G06F 16/3325 Reformulation based on results of preceding query
• CPC Class Codes: G06F 16/95 Retrieval from the web
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising
• CPC Class Codes: G06Q 30/0625 Directed, with specific intent or strategy

Abstract

Query suggestions are provided using a query log including a number of user sessions that comprise training data. The training data includes a sequence of a plurality of sets of queries, some of the sets of queries including query transitions followed by a purchase related event. From a cleaned and normalized query log stationary scores and transition scores of at least some of the plurality of sets are generated. A set of query suggestions is built and similarity scores are computed for at least some of the set of query suggestions to determine whether individual ones of the at least some of the set of query suggestions meet a predetermined assurance level. Those that meet the assurance level are included as elements of the set of query suggestions. The set of query suggestions is mixed and ranked according to a user behavior that is sought to be influenced.

Claims

1. A method for providing query suggestions for a user comprising:

receiving a query;

based on the user query, generating by at least one computer processor, from analysis of a query log of user query sessions on a publication system wherein the user query sessions comprise a plurality of high frequency queries and low frequency queries, query transitions and purchase-related events, a set of transition scores comprising transition scores for ordered pairs of the plurality of high frequency queries and low frequency queries based on transition from a query of an ordered pair of queries to a purchase-related event of the ordered pair of queries;

building a set of query suggestions from the transition scores by a score accumulation process;

determining whether the transition score for at least one member of the set of high frequency query suggestions meets a predetermined confidence threshold; and

providing to a client machine, in response to the received user query, a list of high frequency query suggestions that have transition scores that meet the predetermined confidence threshold;

generating, from the analysis, similarity metrics for suggestions for low frequency queries;

determining whether the similarity metric for a suggestion for a low frequency query meets a predetermined similarity metric; and

providing to a client machine, in response to the received user query, a list of low frequency query suggestions having similarity metrics that meet the predetermined similarity metric but do not meet the predetermined confidence threshold.

2. The method of claim 1 wherein the transition scores comprise a popularity score and a purchase efficiency score and the score accumulation process comprises accumulating the popularity score and accumulating the purchase efficiency score.

3. The method of claim 1 further comprising mixing and ranking the set of query suggestions in accordance with a user behavior that is to be influenced.

4. The method of claim 3 wherein the mixing comprises one of equal mixing and natural mixing.

5. The method of claim 3 wherein the ranking comprises using a first weighting for the popularity score and a second weighting for the purchase efficiency score.

6. The method of claim 1 further comprising cleaning and normalizing the query suggestions.

7. The method of claim 6 wherein the cleaning comprises noise removal and the normalizing comprises dividing the transition scores by the number of users that practice the transition.

8. One or more computer-readable hardware storage device having embedded therein a set of instructions which, when executed by one or more processors of a computer, causes the computer to execute operations comprising:

receiving a user query;

obtaining analysis of a query log of user query sessions on a publication system, wherein the user query sessions comprise a plurality of high frequency queries and low frequency queries, query transitions and purchase-related events;

based on the user query, generating by at least one computer processor, from the analysis of the query log, a set of transition scores comprising transition scores for ordered pairs of high frequency queries and low frequency queries, based on transition from a query of an ordered pair of queries to a purchase-related event of the ordered pair of queries;

building a set of query suggestions for high frequency queries from the transition scores by a score accumulation process;

determining whether the transition score for at least one member of the set of query suggestions meets a predetermined confidence threshold;

providing to a client machine, in response to the received user query, a list of query suggestions for high frequency queries that have transition scores that meet the predetermined confidence threshold;

generating, from the analysis, similarity metrics for suggestions for low frequency queries;

determining whether the similarity metric for a suggestion for a low frequency query meets a predetermined similarity metric; and

providing to a client machine, in response to the received user query, a list of low frequency query suggestions having metrics that meet the predetermined similarity metric but do not meet the predetermined confidence threshold.

9. The one or more hardware storage device of claim 8 wherein the transition scores comprise a popularity score and a purchase efficiency score and the score accumulation process comprises accumulating the popularity score and accumulating the purchase efficiency score.

10. The one or more hardware storage device of claim 8, the operations further comprising mixing and ranking the set of query suggestions in accordance with a user behavior that is to be influenced.

11. The one or more hardware storage device of claim 10 wherein the mixing comprises one of equal mixing and natural mixing.

12. The one or more hardware storage device of claim 10 wherein the ranking comprises using a first weighting for the popularity score and a second weighting for the purchase efficiency score.

13. The one or more hardware storage device of claim 8, the operations further comprising cleaning and normalizing the query suggestions.

14. The one or more hardware storage device of claim 13 wherein the cleaning comprises noise removal and the normalizing comprises dividing the transition scores by the number of users that practice the transition.

15. A system comprising one or more hardware processors configured to include:

a query detection subsystem to detect a user query;

a logging module to receive a query log of client machine sessions on a publication system, the query log comprising high frequency queries and low frequency queries, at least one session comprising a plurality of queries, query transitions and purchase-related events;

a scoring module to generate, responsive to the user query and based on analysis of the query log, transition scores comprising scores for ordered pairs of the high frequency queries and low frequency queries, and further based on transition from a query of an ordered pair of queries to a purchase-related event of the ordered pair of queries, the transition scores comprising a popularity score;

a query suggestion build module to build a set of query suggestions from the transition scores by a score accumulation process;

the scoring module further to compute transition scores for the set of query suggestions to determine whether transition scores for at least one member of the set of query suggestions meets a predetermined confidence threshold;

a similarity scoring module to compute similarity metrics for determination of whether similarity scores meet a predetermined similarity metric; and

an output provider module for providing to a client machine, in response to the detected user query, a list of query suggestions for low frequency queries that have similarity metrics that meet a predetermined similarity metric but do meet the predetermined confidence threshold.

16. The system of claim 15, the one or more hardware processors further configured to include a recommendation compilation module to mix, and a ranking module to rank, the set of query suggestions in accordance with a user behavior that is to be influenced.

17. The system of claim 16 wherein the recommendation compilation module performs one of equal mixing and natural mixing.

18. The system of claim 16 wherein the transition scores further comprise a purchase efficiency score, and the ranking module uses a first weighting for the popularity score and a second weighting for the purchase efficiency score.

19. The system of claim 15, the one or more hardware processors further configured to include a query log cleaning and normalization module to clean the query log and to normalize the query log.

20. The system of claim 19 wherein the cleaning comprises noise removal and the normalizing comprises dividing the transition scores by the number of users that practice the transition.