Integration of multiple query revision models

• US 7,565,345 B2
• Filed: 03/29/2005
• Issued: 07/21/2009
• Est. Priority Date: 03/29/2005
• Status: Active Grant

• Current Assignee: Google LLC (Alphabet Inc.)
• Original Assignee: Google Inc. (Alphabet Inc.)
• Inventors: Bailey, David R., Battle, Alexis J., Nayak, P. Pandurang, Gomes, Benedict A.
• Primary Examiner(s): Rones; Charles
• Assistant Examiner(s): Choi; Yuk Ting
• Application Number:
  US11/094,814
• Publication Number:
  US 20060230022A1
• Time in Patent Office: 1,575 Days
• Field of Search: None
• US Class Current: 1/1
• 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: G06F 16/9532 Query formulation
• CPC Class Codes: Y10S 707/99933 Query processing, i.e. searching
• CPC Class Codes: Y10S 707/99934 Query formulation, input preparation, or translation

Abstract

An information retrieval system includes a query revision architecture that integrates multiple different query revisers, each implementing one or more query revision strategies. A revision server receives a user'"'"'s query, and interfaces with the various query revisers, each of which generates one or more potential revised queries. The revision server evaluates the potential revised queries, and selects one or more of them to provide to the user.

Claims

1. A computer-implemented method comprising:

generating revised queries for an initial query using differing query revision strategies;

calculating a confidence measure for each revised query based on a frequency of occurrence of a query pair comprising the revised query and the initial query;

comparing the confidence measures amongst the generated revised queries;

automatically selecting a subset of the revised queries based on the comparison, the subset of the revised queries including fewer than all of the revised queries;

evaluating search results associated with each revised query of the subset of the revised queries;

automatically selecting a subset of the evaluated search results associated with each revised query of the subset of the revised queries, the subset of the evaluated search results including, for each of the revised queries, fewer than all of the evaluated search results;

storing the subset of the evaluated search results on a memory;

providing, on a first web page, initial search results associated with the initial query; and

providing, on a second web page, a simultaneous display of a plurality of revised queries from the subset of the revised queries and, with each of the plurality of revised queries, one or more evaluated search results associated with the respective query.

2. The method of claim 1, wherein the initial query is received from a front-end server as entered by a client.

3. The method of claim 2, wherein results of the initial query are returned to the client by the front-end server.

4. The method of claim 1, wherein automatically selecting the subset of the revised queries comprises:

sorting the revised queries by the confidence measure to create a ranking of the revised queries.

5. The method of claim 4, further comprising:

obtaining the search results for the subset of the revised queries; and

selecting the subset of the revised queries by evaluating the search results.

6. The method of claim 5, wherein selecting the subset of the revised queries comprises selecting three more of the revised queries wherein:

the search results for the three or more revised queries includes a minimum number of search results;

the search results for the three or more revised queries includes a minimum number of new top results; and

the three or more revised queries do not exceed a predetermined maximum number of the revised queries.

7. The method of claim 6, wherein the minimum number of search results is 1, the minimum number of new top results is 2, and the predetermined maximum number of the selected revised queries is 4.

8. The method of claim 5, wherein obtaining the search results for the subset of the revised queries comprises:

providing the subset of the revised queries to a search engine; and

receiving the results of the subset of the revised queries from the search engine.

9. The method of claim 1, wherein the differing query revision strategies include strategies for broadening at least one query term, refining at least one query term, syntactically revising at least one query term, and session-based revising of at least one query term.

13. The method of claim 1, wherein automatically selecting the subset of the revised queries further comprises excluding at least one of the generated revised queries.

14. The method of claim 1, further comprising:

generating a first table of recurring individual queries from user session query data, the recurring individual queries including a first query;

generating a second table of recurring query pairs from the user session query data, the recurring query pairs including the first query and a second query that followed the first query;

determining a first occurrence count of the initial query in the first table;

determining a second occurrence count of the query pair in the second table; and

calculating the frequency of occurrence based on dividing the second occurrence count by the first occurrence count.

18. The method of claim 1, further comprising:

providing, with the initial search results, a link to the second web page; and

receiving an indication of a user selection of the link.

20. The method of claim 1, wherein the operations further comprise:

providing, with the initial search results, a link to the second web page; and

receiving an indication of a user selection of the link.

21. The method of claim 1, wherein the subset of the evaluated search results are displayed below each of the subset of the revised queries.

24. The method of claim 1, further comprising:

ranking each of the revised queries of the subset of the revised queries;

selecting a top three ranked revised queries based on the ranking; and

for each of the top three ranked revised queries;

ranking each of the one or evaluated search results of the subset of the evaluated search results associated with the respective top three ranked query; and

selecting a top three ranked evaluated search results based on the ranking.

27. The method of claim 1, wherein the initial query represents a previously revised query.

10. A computer-readable storage medium encoded with a computer program comprising instructions that, when executed, operate to cause a computer to perform operations comprising:

generating revised queries for an initial query using differing query revision strategies;

calculating a confidence measure for each revised query based on a frequency of occurrence of a query pair comprising the revised query and the initial query;

comparing the confidence measures amongst the generated revised queries;

automatically selecting a subset of the revised queries based on the comparison, the subset of the revised queries including fewer than all of the revised queries;

evaluating search results associated with each query of the subset of the revised queries;

automatically selecting a subset of the evaluated search results associated with each revised query of the subset of the revised queries, the subset of the evaluated search results including, for each of the revised queries, fewer than all of the evaluated search results;

storing the subset of the evaluated search results on a memory;

providing, on a first web page, initial search results associated with the initial query;

providing, on a second web page, a simultaneous display of a plurality of revised queries from the subset of the revised queries and, with each of the plurality of revised queries, one or more evaluated search results associated with the respective query.

15. The computer-readable medium of claim 10, wherein automatically selecting the subset of the revised queries further comprises excluding at least one of the generated revised queries.

16. The computer-readable medium of claim 10, wherein generating the revised queries comprises:

providing the initial query to revisers implementing the differing query revision strategies; and

receiving from at least one of the revisers two or more of the revised queries.

19. The computer-readable medium of claim 10, wherein the operations further comprise:

providing, with the initial search results, a link to the second web page; and

receiving an indication of a user selection of the link.

22. The computer-readable medium of claim 10, wherein the subset of the evaluated search results are displayed below each of the subset of the revised queries.

25. The computer-readable medium of claim 10, wherein the operations further comprise:

ranking each of the revised queries of the subset of the revised queries;

selecting a top three ranked revised queries based on the ranking; and

for each of the top three ranked revised queries;

ranking each of the one or evaluated search results of the subset of the evaluated search results associated with the respective top three ranked query; and

selecting a top three ranked evaluated search results based on the ranking.

28. The computer-readable medium of claim 10, wherein the initial query represents a previously revised query.

11. A system, comprising:

one or more computers, and one or more processors configured to perform operations comprising;

generating revised queries for an initial query using differing query revision strategies;

calculating a confidence measure for each revised query based on a frequency of occurrence of a query pair comprising the revised query and the initial query;

comparing the confidence measures amongst the generated revised queries;

automatically selecting a subset of the revised queries based on the comparison, the subset of the revised queries including fewer than all of the revised queries;

evaluating search results associated with each query of the subset of the revised queries;

automatically selecting a subset of the evaluated search results associated with each revised query of the subset of the revised queries, the subset of the evaluated search results including, for each of the revised queries, fewer than all of the evaluated search results;

providing, on a first web page, initial search results associated with the initial query; and

providing, on a second web page, a simultaneous display of a plurality of revised queries from the subset of the revised queries and, with each of the plurality of revised queries, one or more evaluated search results associated with the respective query.

12. The system of claim 11, wherein the system is customizable for specific subject matter domains.

17. The system of claim 11, wherein generating the revised queries comprises:

providing the initial query to revisers implementing the differing query revision strategies; and

receiving from at least one of the revisers two or more of the revised queries.

23. The system of claim 11, wherein the subset of the evaluated search results are displayed below each of the subset of the revised queries.

26. The system of claim 11, wherein the operations further comprise:

ranking each of the revised queries of the subset of the revised queries;

selecting a top three ranked revised queries based on the ranking; and

for each of the top three ranked revised queries;

ranking each of the one or evaluated search results of the subset of the evaluated search results associated with the respective top three ranked query; and

selecting a top three ranked evaluated search results based on the ranking.

29. The system of claim 11, wherein the initial query represents a previously revised query.