Selection of search phrases to suggest to users in view of actions performed by prior users

• US 20040236736A1
• Filed: 06/22/2004
• Published: 11/25/2004
• Est. Priority Date: 12/10/1999
• Status: Active Grant

• Current Assignee: A9.com Incorporated (Amazon.com, Inc.)
• Original Assignee: A9.com Incorporated (Amazon.com, Inc.)
• Inventors: Scofield, Christopher L., Whitman, Ronald M.
• Granted Patent: US 7,424,486 B2
• Time in Patent Office: Days
• US Class Current: 707/3
• CPC Class Codes: G06F 16/24534 Query rewriting; Transformation
• CPC Class Codes: G06F 16/3322 using system suggestions G06F16/3325 takes precedence
• CPC Class Codes: G06F 16/951 Indexing; Web crawling techniques
• CPC Class Codes: Y10S 707/99932 Access augmentation or optimizing
• CPC Class Codes: Y10S 707/99933 Query processing, i.e. searching
• CPC Class Codes: Y10S 707/99935 Query augmenting and refining, e.g. inexact access
• CPC Class Codes: Y10S 707/99936 Pattern matching access
• CPC Class Codes: Y10S 707/99942 Manipulating data structure, e.g. compression, compaction, compilation
• CPC Class Codes: Y10S 707/99943 Generating database or data structure, e.g. via user interface
• CPC Class Codes: Y10S 707/99944 Object-oriented database structure
• CPC Class Codes: Y10S 707/99945 Object-oriented database structure processing

Abstract

A search engine system uses information about historical query submissions to a search engine to suggest previously-submitted, related search phrases to users. The related search phrases are preferably suggested based on a most recent set of query submission data (e.g., the last two weeks of submissions), and thus strongly reflect the current searching patterns or interests of users. In one embodiment, the related search phrases are scored and selected for display based at least in-part on (a) a frequency with which each search phrase has been submitted, and/or (b) an evaluation of the “usefulness” of each search phrase, as reflected by actions performed by prior users while viewing corresponding search results.

Claims

1. A search engine system, comprising:

a search phrase selection component that selects search phrases from a record of prior search query submissions of a plurality of users of a search engine, each search phrase comprising a sequence of two or more terms;

a data structure that associates each of a plurality of key terms with a corresponding set of related search phrases selected by the search phrase selection component; and

a query processing component that is responsive to a submission by a user of a single-term search query containing a key term by looking up from the data structure a corresponding set of related search phases to suggest to the user.

2. The search engine system of claim 1, wherein the search phrase selection component selects search phrases based at least in part on frequencies with which specific search phrases have been submitted to the search engine, as reflected within the record of prior search query submissions.

3. The search engine system of claim 1, wherein the search phrase selection component evaluates a candidate search phrase based at least in part on a frequency with which users who have submitted the candidate search phrase to perform a search have thereafter selected a corresponding search result item for viewing.

4. The search engine system of claim 1, wherein the search phrase selection component evaluates a candidate search phrase at least in part by analyzing post-search actions performed by users who submitted the candidate search phrase.

5. The search engine system of claim 1, wherein the search phrase selection component implements a scoring algorithm for generating scores for specific search phrases, and uses the scores to select search phrases to include in the data structure.

6. The search engine system of claim 1, wherein the data structure maps each key term to a set of one or more search phrases that contain the respective key term.

7. The search engine system of claim 1, wherein the query processing component additionally uses the data structure to look up search phrases to suggest in response to multiple-term search queries.

8. A method of assisting users of a search engine in refining searches, the method comprising:

programmatically analyzing data reflective of searches conducted by a plurality of users over a period of time to select, for a key term, a set of previously-submitted search phrases that are related to the key term;

generating a data structure that associates the key term with the set of related search phrases;

receiving a search query specified by a user to search a repository of items, wherein the search query includes the key term; and

looking up from the data structure at least one of the related search phrases to suggest to the user.

9. The method of claim 8, wherein a candidate search phrase is considered for inclusion in said set based at least in part on a frequency with which the candidate search phrase has been submitted over said period of time.

10. The method of claim 8, wherein a candidate search phrase is considered for inclusion in said set based at least in part on a frequency with which users who have submitted the candidate search phrase to perform a search have thereafter selected a corresponding search result item for viewing.

11. The method of claim 8, wherein a candidate search phrase is considered for inclusion in said set based at least in part on results of an automated analysis of post-search actions performed by users who submitted the candidate search phrase.

12. The method of claim 8, wherein programmatically analyzing the data reflective of searches comprises identifying a plurality of previously submitted search phrases that contain the key term, generating scores for each of the previously submitted search phrases, and using the scores to select said set of search phrases.

13. The method of claim 8, further comprising presenting the at least one related search phrase to the user within a search results page for said search query.

14. The method of claim 8, wherein the search query specified by the user is a single-term search query.

15. The method of claim 8, wherein each search phrase is an ordered set of two or more terms.

16. The method of claim 8, wherein each search phrase in the set includes the key term and at least one additional term.

17. The method of claim 8, wherein the key term and the set of related search phrases are stored in the data structure in association with a search context.

18. A computer system that performs the method of claim 8.

19. A computer-readable medium having stored thereon a computer program that, when executed, performs the method of claim 8.

20. A method of processing search queries submitted by users to search a repository of items, the method comprising:

receiving a user submission of a single-term search query that contains a single search term;

executing the search query to identify items that are responsive to the search query;

accessing a data structure to look up a previously-submitted search phrase that includes the search term and that, based at least in-part on a computer analysis of prior search query submissions of a plurality of users, has been ranked highly relative to other previously-submitted search phrases that include the search term, said search phrase comprising a sequence of two or more search terms; and

generating a search results page that lists at least some of the responsive items and provides a user option to select the search phrase to submit a refined search query.

21. The method of claim 20, wherein the ranking of the search phrase reflects a frequency with which the search phrase has been submitted by users to conduct searches.

22. The method of claim 20, wherein the ranking of the search phrase reflects a usefulness of the search phrase as measured by analyzing post-search actions of users who submitted the search phrase.

23. The method of claim 20, wherein the ranking of the search phrase reflects a frequency with which users who submitted the search phrase thereafter selected a resulting search result item.

24. The method of claim 20, further comprising ranking the search phrase based on at least one of the following:

(a) a frequency with which the search phrase has been submitted by users to conduct searches, (b) a frequency with which users who submitted the search phrase thereafter selected a resulting search result item.

25. The method of claim 20, wherein the search query is a field-restricted search query that specifies a search field, and wherein the search field and the search term are used in combination to look up the previously-submitted search phrase from the data structure.

26. The method of claim 20, wherein the method comprises looking up from the data structure, and incorporating into the search results page, a plurality of previously-submitted search phrases that include the search term.

27. A computer system that performs the method of claim 20.

28. A computer-readable medium having stored thereon a computer program that, when executed, performs the method of claim 20.

29. A computer-implemented method of identifying search phrases to suggest to users, the method comprising

monitoring and recording transactional events associated with searches conducted by users of a search engine to generate search event data;

identifying within the search event data a search phrase submitted by users to conduct searches, said search phrase comprising a sequence of two or more search terms; and

programmatically analyzing the search event data associated with user submissions of the search phrase to generate a score that represents a desirability of suggesting the search phrase to users.

30. The method of claim 29, further comprising comparing the score to respective scores of other search phrases that share a common key term with the search phrase to select a set of search phrases to suggest to users who use the key term to conduct searches.

31. The method of claim 30, further comprising mapping the key term to the selected set of search phrases in a data structure, and using the data structure to look up search phrases to suggest in response to search query submission events.

32. The method of claim 29, further comprising using the score to rank the search phrase relative to other search phrases that share a common search term with the search phrase.

33. The method of claim 29, wherein analyzing the search event data to generate the score comprises taking into consideration a frequency with which users have submitted the search phrase to conduct searches.

34. The method of claim 29, wherein analyzing the search event data to generate the score comprises taking into consideration a frequency with which users who submitted the search phrase thereafter selected a resulting search result item.

35. The method of claim 29, wherein analyzing the search event data to generate the score comprises taking into consideration post-query-submission user actions that reflect a usefulness of the search phrase to users.

36. The method of claim 29, wherein the score reflects a usefulness of the search phrase to users as measured by analyzing post-search actions of users who submitted the search phrase to conduct a search.

37. The method of claim 29, wherein the score additionally reflects a number of search result items returned when the search phrase is used to conduct a search.

38. The method of claim 29, further comprising using the score to determine whether to suggest the search phrase to users that submit search queries that are related to the search phrase.

39. The method of claim 29, further comprising using the score to determine whether to suggest the search phrase to users that submit a single-term search query containing a single search term of said search phrase.

40. A computer system that performs the method of claim 29.

41. A computer-readable medium having stored thereon a computer program that, when executed, performs the method of claim 29.