Culturally relevant search results

• US 7,624,103 B2
• Filed: 10/04/2006
• Issued: 11/24/2009
• Est. Priority Date: 07/21/2006
• Status: Active Grant

• Current Assignee: Meta Platforms, Inc. (f/k/a Facebook, Inc.)
• Original Assignee: AOL LLC (Apollo Global Management, Inc.)
• Inventors: Vandermolen, Harmannus, Sommers, Michael, Howe, Karen, Wiegering, Anthony
• Primary Examiner(s): LEROUX, ETIENNE PIERRE
• Application Number:
  US11/538,729
• Publication Number:
  US 20080021860A1
• Time in Patent Office: 1,147 Days
• Field of Search: 707/2, 707/3, 707/4, 705/36
• US Class Current: 1/1
• CPC Class Codes: G06F 16/24578 using ranking
• CPC Class Codes: G06F 16/951 Indexing; Web crawling techniques
• CPC Class Codes: G06F 16/9535 Search customisation based on user profiles and personalisation
• CPC Class Codes: G06F 16/9538 Presentation of query results
• CPC Class Codes: Y10S 707/99934 Query formulation, input preparation, or translation

Abstract

Search results may be provided to a user. A search query may be received from the user. A query feature vector may be formed for the search query. The query feature vector may be compared with news feature vectors associated with documents related to current events. An augmented query feature vector may be formed based on results of the comparison of the query feature vector with the news feature vectors. The augmented query feature vector may be compared with feature vectors related to target documents. Search results that include target documents may be identified based on results of the comparison of the augmented query feature vector with the feature vectors related to the target documents. The user may be made able to perceive at least some of the identified search results.

Claims

1. A method for providing search results to a user, the method comprising:

receiving a search query submitted by the user to a search service, the search query being submitted by the user on a particular date;

forming, using at least one computer processor, a query feature vector for the search query, the query feature vector being represented as a set of numerical values that include numerical values representing the terms submitted in the search query;

accessing news feature vectors associated with documents related to current events, the documents related to current events being documents having a publication date determined to be temporally proximate to the particular date of submission of the search query and each of the news feature vectors being represented as a set of numerical values that includes numerical values representing words included in a document related to a current event;

comparing the query feature vector with the news feature vectors by calculating distances or angles between the query feature vector and the news feature vectors;

forming an augmented query feature vector based on results of the comparison of the query feature vector with the news feature vectors;

accessing target feature vectors related to target documents, each target feature vector being represented as a set of numerical values that include numerical values representing words included in a target document;

comparing the augmented query feature vector with the target feature vectors by calculating distances or angles between the augmented query feature vector and the target feature vectors;

identifying search results that include target documents based on results of the comparison of the augmented query feature vector with the target feature vectors; and

enabling the user to perceive at least some of the identified search results.

2. The method of claim 1 wherein enabling the user to perceive the identified search results comprises enabling the user to perceive only the identified search results.

3. The method of claim 2 further comprising:

determining relevance of each of the identified search results to the search query; and

sorting the identified search results based on the determined relevance.

4. The method of claim 1 wherein enabling the user to perceive the search results comprises:

sorting the identified search results into a first group of search results including identified search results that are closely related to a current event and a second group including identified search results that have no particular known relationship with the current event; and

enabling the user to perceive a difference between the first and second groups of identified search results.

5. The method of claim 1 further comprising enabling the user to perceive other search results in addition to and integrated with the identified search results.

6. The method of claim 5 further comprising:

determining relevance of each identified and other search result to the search query; and

sorting the identified and other search results based on the determined relevance.

7. The method of claim 1 further comprising:

determining news result feature vectors based on results of the comparison of the query feature vector with the news feature vectors; and

determining a centroid feature vector based on the news result feature vectors, wherein the centroid feature vector is represented as a new set of numerical values determined by calculating a weighted average of the sets of numerical values that represent the news result feature vectors,

wherein forming the augmented query feature vector comprises forming the augmented query feature vector based on the centroid feature vector.

8. The method of claim 1 further comprising

determining news result feature vectors based on results of the comparison of the query feature vector with the news feature vectors;

and

determining a centroid feature vector based on the news result feature vectors, wherein the centroid feature vector is represented as a new set of numerical values determined based on a clustering analysis of the news feature vectors,

wherein forming the augmented query feature vector comprises forming the augmented query feature vector based on the centroid feature vector.

9. The method of claim 1 further comprising:

accessing a target corpus data store that references target documents, wherein target documents include at least some documents related to current events; and

forming a target feature vector for each target document.

10. The method of claim 1 further comprising:

accessing a news corpus data store that references documents related to current events, wherein the current events are events related to human action; and

forming a news feature vector for each document related to the current events.

11. The method of claim 1 wherein enabling the user to perceive the search results comprises:

determining a relationship between the search results and a current event; and

providing the search results to the user as a perceivable list of search results organized with search results that are closely related to the current event being positioned closer to the top of the list than other search results.

12. The method of claim 1 wherein enabling the user to perceive the search results comprises:

distinguishing search results that are closely related to a current event from search results that have no particular known relationship with the current event; and

providing the search results to the user as a perceivable list of search results organized with the search results that are closely related to the current event being separated from the search results that have no particular known relationship with the current event using a label.

13. The method of claim 1 wherein enabling the user to perceive the search results comprises:

determining search results that are closely related to a current event;

determining search results that have no particular known relationship with the current event;

blending the search results that are closely related to the current event with the search results that have no particular known relationship with the current event; and

providing the search results to the user as a perceivable list of search results organized with the search results that are closely related to the current event being indistinguishable from the search results that have no particular known relationship with the current event.

14. The method of claim 1 wherein enabling the user to perceive the search results comprises:

determining search results that are closely related to a current event;

determining search results that have no particular known relationship with the current event;

blending the search results that are closely related to the current event with the search results that have no particular known relationship with the current event; and

providing the search results to the user as a perceivable list of search results organized with the search results that are closely related to the current event being distinguishable from the search results that have no particular known relationship with the current event.

15. The method of claim 1, wherein comparing the query feature vector with the news feature vectors includes:

calculating, using the set of numerical values of the query feature vector and the set of numerical values of the news feature vectors as coordinates, the euclidean distances or angles between the query feature vector and the news feature vectors to identify news feature vectors that are within a predetermined euclidean distance or angle of the query feature vector; and

designating the identified news feature vectors as news result feature vectors.

16. The method of claim 15 wherein forming the augmented query feature vector comprises:

identifying terms that are included in one or more of the news result feature vectors, but absent from the query feature vector; and

representing the augmented query feature vector as a new set of numerical values that includes numerical values that represent the identified terms and includes the set of numerical values that represents the query feature vector.

17. The method of claim 15 wherein forming the augmented query feature vector comprises:

identifying terms included in the query feature vector, but absent from the news result feature vectors; and

representing the augmented query feature vector as the set of numerical values that represents the query feature vector but excluding numerical values that represent the identified terms from the set of numerical values that represents the query feature vector.

18. The method of claim 15 wherein forming the augmented query feature vector comprises:

identifying terms included in the query feature vector that are similar to, but different from, terms included in the news result feature vectors; and

representing the augmented query feature vector as the set of numerical values that represents the query feature vector but modifying the numerical values representing the identified terms to conform to the identified terms in the news result feature vectors.

19. The method of claim 15, wherein calculating comprises calculating euclidean distances between a point in n-dimensional space represented by the set of numerical values of the query feature vector and points in n-dimensional space represented by the sets of numerical values of the news feature vectors.

20. The method of claim 15, wherein calculating comprises calculating angles between a vector that extends to a point in n-dimensional space represented by the set of numerical values of the query feature vector and vectors that extend to points in n-dimensional space represented by the sets of numerical values of the news feature vectors.

21. The method of claim 1, wherein:

comparing the augmented query feature vector with the target feature vectors includes calculating, using the set of numerical values of the augmented query feature vector and the sets of numerical values of the target feature vectors as coordinates, the euclidean distances or angles between the augmented query feature vector and the target feature vectors to identify target feature vectors that are within a predetermined euclidean distance or angle of the augmented query feature vector, and

identifying the search results comprises designating the identified target feature vectors as the identified search results.

22. The method of claim 21, wherein calculating comprises calculating euclidean distances between a point in n-dimensional space represented by the set of numerical values of the augmented query feature vector and points in n-dimensional space represented by the sets of numerical values of the target feature vectors.

23. The method of claim 21, wherein calculating comprises calculating angles between a vector that extends to a point in n-dimensional space represented by the set of numerical values of the augmented query feature vector and vectors that extend to points in n-dimensional space represented by the sets of numerical values of the target feature vectors.

24. The method of claim 1, wherein the set of numerical values of the query feature vector includes a single numerical value for each term submitted in the search query.

25. The method of claim 1, wherein each news feature vector corresponds to a single document related to a current event and each target feature vector corresponds to a single target document.

26. The method of claim 1, wherein a document related to a current event is determined to be temporally proximate to the particular date of submission of the search query if the document was published by a news outlet less than or equal to 24 hours prior to submission of the search query by the user.

27. The method of claim 1, wherein a document related to a current event is determined to be temporally proximate to the particular date of submission of the search query if the document is associated with science and was published less than or equal to 6 months prior to submission of the search query by the user.

28. The method of claim 1, wherein the documents are web pages.

29. A computer program product for providing search results to a user, the computer program product including instructions that, when executed, cause a processor to:

receive a search query submitted by the user to a search service, the search query being submitted by the user on a particular date;

form a query feature vector for the search query, the query feature vector being represented as a set of numerical values that include numerical values representing the terms submitted in the search query;

access news feature vectors associated with documents related to current events, the documents related to current events being documents having a publication date determined to be temporally proximate to the particular date of submission of the search query and each of the news feature vectors being represented as a set of numerical values that includes numerical values representing words included in a document related to a current event;

compare the query feature vector with the news feature vectors by calculating distances or angles between the query feature vector and the news feature vectors;

form an augmented query feature vector based on results of the comparison of the query feature vector with the news feature vectors;

access target feature vectors related to target documents, each target feature vector being represented as a set of numerical values that include numerical values representing words included in a target document;

compare the augmented query feature vector with the target feature vectors by calculating distances or angles between the augmented query feature vector and the target feature vectors;

identify search results that include target documents based on results of the comparison of the augmented query feature vector with the target feature vectors; and

enable the user to perceive at least some of the identified search results.

30. A system for providing search results to a user, the system comprising:

means for receiving a search query submitted by the user to a search service, the search query being submitted by the user on a particular date;

means for forming, using at least one computer processor, a query feature vector for the search query, the query feature vector being represented as a set of numerical values that include numerical values representing the terms submitted in the search query;

means for accessing news feature vectors associated with documents related to current events, the documents related to current events being documents having a publication date determined to be temporally proximate to the particular date of submission of the search query and each of the news feature vectors being represented as a set of numerical values that includes numerical values representing words included in a document related to a current event;

means for comparing the query feature vector with the news feature vectors by calculating distances or angles between the query feature vector and the news feature vectors;

means for forming an augmented query feature vector based on results of the comparison of the query feature vector with the news feature vectors;

means for accessing target feature vectors related to target documents, each target feature vector being represented as a set of numerical values that include numerical values representing words included in a target document;

means for comparing the augmented query feature vector with the target feature vectors by calculating distances or angles between the augmented query feature vector and the target feature vectors;

means for identifying search results that include target documents based on results of the comparison of the augmented query feature vector with the target feature vectors; and

means for enabling the user to perceive at least some of the identified search results.