Filtering search results

• US 7,571,157 B2
• Filed: 12/29/2004
• Issued: 08/04/2009
• Est. Priority Date: 12/29/2004
• Status: Expired due to Fees

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: AOL LLC (Apollo Global Management, Inc.)
• Inventors: Pass, Gregory S., Chowdhury, Abdur R.
• Primary Examiner(s): Rones; Charles
• Assistant Examiner(s): Lewis; Alicia M
• Application Number:
  US11/023,633
• Publication Number:
  US 20060143159A1
• Time in Patent Office: 1,679 Days
• Field of Search: None
• US Class Current: 1/1
• CPC Class Codes: G06F 16/33 Querying
• 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 707/99937 Sorting

Abstract

Search results may be sorted or filtered based on scores assigned to the search results. For example, scores may be assigned to search results based on characteristics of surrogate representations of the search results, which are relatively short summaries or excerpts of the search result that may be presented in place of the search results themselves, and those assigned scores may be used to sort or filter the search results. In one example, pairs of search results may be examined to identify significant drops in quality between search results, which is indicated by a large relative or absolute difference in the scores of the search results. Search results with scores that indicate ranks that are lower than a lower ranked search result of the pair of search results may be eliminated when the difference between the scores assigned to the pair of search results exceeds a maximum allowable difference.

Claims

1. A computer-implemented method for filtering a set of search results, the method comprising:

retrieving a set of search results for a query received from a user;

assigning a relevancy score to each of the retrieved search results based on relevancy of the retrieved search result to the query;

selecting, from the set of search results, two adjacent search results based on relevancy scores assigned thereto, wherein the two adjacent search results are selected such that no other search results have assigned relevancy scores that are between the relevancy scores assigned to the two adjacent search results, the two adjacent search results include a first search result that is assigned a first relevancy score and a second result that is assigned a second relevancy score that is different than the first relevancy score, and the first and second relevancy scores indicate that the second search result is less relevant to the query than the first search result;

determining a difference between the relevancy scores assigned to the two adjacent search results, the difference reflecting a difference between the relevancy of the first search result to the query and the relevancy of the second search result to the query;

accessing data reflecting a maximum allowable difference in relevancy;

determining that the difference between the relevancy scores assigned to the two adjacent search results exceeds the accessed maximum allowable difference;

in response to the determination that the difference between the relevancy scores assigned to the two adjacent search results exceeds the maximum allowable difference, eliminating, from the set of search results, the search results with assigned relevancy scores that indicate a relevancy to the query that is less than or equal to the relevancy of the second search result to the query; and

storing, in electronic storage, at least one of the search results that have not been eliminated.

2. The method of claim 1 further comprising eliminating the search results with assigned scores that are less than or equal to a minimum allowable score.

3. The method of claim 1 wherein assigning a score to each of the retrieved search results comprises assigning to the search result a score that was calculated by a source from which the search result was retrieved.

4. The method of claim 1 wherein determining a difference between the scores assigned to the two search results comprises determining an absolute difference between the scores assigned to the two search results.

5. The method of claim 1 wherein determining a difference between the scores assigned to the two search results comprises determining a difference between the scores assigned to the two search results that is relative to at least one score of one of the search results.

6. The method of claim 1 wherein:

retrieving search results comprises retrieving surrogate representations of the search results; and

assigning a score to each of the retrieved search results comprises assigning a score to each of the retrieved search results based on visual characteristics of the surrogate representations corresponding thereto.

7. The method of claim 1 further comprising sorting the search results based on the assigned scores.

8. The method of claim 1 wherein retrieving search results for a query received from a user comprises:

submitting the received query to a search engine; and

receiving search results for the received query from the search engine.

9. The method of claim 1 wherein retrieving search results for a query received from a user comprises:

submitting the received query to multiple search engines; and

receiving search results for the received query from each of the multiple search engines.

10. The method of claim 1 further comprising making at least one of the search results that have not been eliminated perceivable to the user.

16. The method of claim 1 wherein the maximum allowable difference is a percentage of a difference between the maximum identified score and the minimum identified score.

17. The method of claim 1 wherein the maximum allowable difference is a percentage of a difference between the scores identified for the two search results.

19. The method of claim 1 wherein eliminating the search results comprises at least one of filtering the search results and eliminating the search results from display.

21. The method of claim 1 further comprising:

dynamically computing the maximum allowable difference based on at least one of the assigned relevancy scores of the retrieved search results.

22. The method of claim 21 wherein dynamically computing the maximum allowable difference based on at least one of the assigned relevancy scores of the retrieved search results comprises dynamically computing the maximum allowable difference based on all of the assigned relevancy scores of the retrieved search results.

23. The method of claim 1 wherein the maximum allowable difference is at least one of a percentage of a maximum identified score, a percentage of a minimum identified score, a percentage of an average identified score, and a percentage of a standard deviation of the identified scores.

25. The method of claim 1 wherein:

retrieving the set of search results for the query received from the user comprises retrieving a single list of search results responsive to receipt of the query;

selecting, from the set of search results, the two adjacent search results comprises selecting, from the single list of search results, two adjacent search results;

eliminating, from the set of search results, the search results with assigned relevancy scores that indicate a relevancy to the query that is less than or equal to the relevancy of the second search result to the query comprises eliminating, from the single list of search results, the search results in the single list with assigned relevancy scores that indicate a relevancy to the query that is less than or equal to the relevancy of the second search result to the query; and

storing, in electronic storage, at least one of the search results that have not been eliminated comprises storing, in electronic storage, the single list of search results after eliminating, from the single list of search results, the search results in the single list with assigned relevancy scores that indicate a relevancy to the query that is less than or equal to the relevancy of the second search result to the query.

26. The method of claim 21 wherein dynamically computing the maximum allowable difference based on at least one of the assigned relevancy scores of the retrieved search results comprises, subsequent to assigning a relevancy score to each of the retrieved search results and prior to determining that the difference between the relevancy scores assigned to the two adjacent search results exceeds the accessed maximum allowable difference, computing the maximum allowable difference using an equation and using at least one of the assigned relevancy scores of the retrieved search results as a variable in evaluating the equation.

11. A computer-implemented method for filtering a set of search results, the method comprising:

retrieving search results for a query received from a user;

assigning a relevancy score to each of the retrieved search results based on relevancy of the retrieved search result to the query;

selecting two of the search results based on relevancy scores assigned thereto, wherein the two search results are selected such that no other search results have assigned relevancy scores that are between the relevancy scores assigned to the two search results, the two search results include a first search result that is assigned a first relevancy score and a second result that is assigned a second relevancy score that is different than the first relevancy score, and the first and second relevancy scores indicate that the second search result is less relevant to the query than the first search result;

determining a difference between the relevancy scores assigned to the two search results, the difference reflecting a difference between the relevancy of the first search result to the query and the relevancy of the second search result to the query;

eliminating the search results with assigned relevancy scores that indicate a relevancy to the query that is less than or equal to the relevancy of the second search result to the query when the difference between the relevancy scores assigned to the two search results exceeds a maximum allowable difference in relevancy to the query; and

storing, in electronic storage, at least one of the search results that have not been eliminated,

wherein the maximum allowable difference is at least one of a percentage of a maximum identified score, a percentage of a minimum identified score, a percentage of an average identified score, and a percentage of a standard deviation of the identified scores.

14. The method of claim 11 wherein the maximum allowable difference is a percentage of a minimum identified score.

15. The method of claim 11 wherein the maximum allowable difference is a percentage of an average identified score.

18. The method of claim 11 wherein the maximum allowable difference is a percentage of a standard deviation of the identified scores.

24. The method of claim 11 wherein the maximum allowable difference is a percentage of a maximum identified score.

12. A machine-accessible storage medium that when accessed, results in a machine performing operations for filtering a set of search results, comprising:

retrieving a set of search results for a query received from a user;

assigning a relevancy score to each of the retrieved search results based on relevancy of the retrieved search result to the query;

selecting, from the set of search results, two adjacent search results based on relevancy scores assigned thereto, wherein the two adjacent search results are selected such that no other search results have assigned relevancy scores that are between the relevancy scores assigned to the two adjacent search results, the two adjacent search results include a first search result that is assigned a first relevancy score and a second result that is assigned a second relevancy score that is different than the first relevancy score, and the first and second relevancy scores indicate that the second search result is less relevant to the query than the first search result;

determining a difference between the relevancy scores assigned to the two adjacent search results, the difference reflecting a difference between the relevancy of the first search result to the query and the relevancy of the second search result to the query;

accessing data reflecting a maximum allowable difference in relevancy;

determining that the difference between the relevancy scores assigned to the two adjacent search results exceeds the accessed maximum allowable difference;

in response to the determination that the difference between the relevancy scores assigned to the two adjacent search results exceeds the maximum allowable difference, eliminating, from the set of search results, the search results with assigned relevancy scores that indicate a relevancy to the query that is less than or equal to the relevancy of the second search result to the query; and

storing, in electronic storage, at least one of the search results that have not been eliminated.

13. A system comprising:

at least one computer; and

at least one computer-readable medium coupled to the at least one computer having instructions stored thereon which, when executed by the at least one computer, causes the at least one computer to perform operations comprising;

retrieving a set of search results for a query received from a user;

assigning a relevancy score to each of the retrieved search results based on relevancy of the retrieved search result to the query;

selecting, from the set of search results, two adjacent search results based on relevancy scores assigned thereto, wherein the two adjacent search results are selected such that no other search results have assigned relevancy scores that are between the relevancy scores assigned to the two adjacent search results, the two adjacent search results include a first search result that is assigned a first relevancy score and a second result that is assigned a second relevancy score that is different than the first relevancy score, and the first and second relevancy scores indicate that the second search result is less relevant to the query than the first search result;

determining a difference between the relevancy scores assigned to the two adjacent search results, the difference reflecting a difference between the relevancy of the first search result to the query and the relevancy of the second search result to the query;

accessing data reflecting a maximum allowable difference in relevancy;

determining that the difference between the relevancy scores assigned to the two adjacent search results exceeds the accessed maximum allowable difference;

in response to the determination that the difference between the relevancy scores assigned to the two adjacent search results exceeds the maximum allowable difference, eliminating, from the set of search results, the search results with assigned relevancy scores that indicate a relevancy to the query that is less than or equal to the relevancy of the second search result to the query; and

storing, in electronic storage, at least one of the search results that have not been eliminated.

20. A computer-implemented method for filtering a set of search results, the method comprising:

retrieving search results for a query received from a user;

assigning a relevancy score to each of the retrieved search results;

identifying a maximum relevancy score of the search results;

identifying a minimum relevancy score of the search results;

computing a difference between the maximum relevancy score and the minimum relevancy score;

selecting two search results based on relevancy scores assigned thereto, wherein the two search results are selected such that no other search results have assigned relevancy scores that are between the relevancy scores assigned to the two search results, the two search results include a first search result that is assigned a first relevancy score and a second result that is assigned a second relevancy score that is different than the first relevancy score, and the first and second relevancy scores indicate that the second search result is less relevant to the query than the first search result;

computing a difference between the first relevancy score and the second relevancy score;

computing a ratio of the difference between the first relevancy score and the second relevancy score with respect to the difference between the maximum relevancy score and the minimum relevancy score;

comparing the computed ratio to a threshold ratio of a difference between relevancy scores with respect to a difference between a maximum identified score and a minimum identified score;

based on comparison results, determining whether the computed ratio exceeds the threshold ratio;

conditioned on determining that the computed ratio exceeds the threshold ratio, eliminating the search results with assigned relevancy scores that indicate a rank that is lower than a rank of the second search result; and

storing, in electronic storage, at least one of the search results that have not been eliminated.