Systems and methods for personalized search

• US 20060248059A1
• Filed: 08/09/2005
• Published: 11/02/2006
• Est. Priority Date: 04/29/2005
• Status: Active Grant

• Current Assignee: Xerox Corporation (Xerox Holdings Corp.)
• Original Assignee: Palo Alto Research Center, Inc. (Xerox Holdings Corp.)
• Inventors: Gasparetti, Fabio, Chi, Ed
• Granted Patent: US 8,606,781 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G06F 16/9535 Search customisation based on user profiles and personalisation
• CPC Class Codes: G06F 16/9536 Search customisation based on social or collaborative filtering
• CPC Class Codes: G06F 16/9538 Presentation of query results

Abstract

Techniques are presented to provide personalized search results to a user. A user history optionally classified into projects or tasks is determined. A profile for the user is determined by identifying keywords, concepts or other user and/or task specific descriptors within user history documents. A proximal neighborhood based on the user'"'"'s history and adjustable crawling parameters is determined. The adjustable crawling parameters define which documents linked to documents in the user history are included within the proximal neighborhood. A user query is determined and used to search the set of documents within the proximal neighborhood. The results are then ranked based on the user profile information. The ranked results are optionally displayed to the user.

Claims

1. A method for personalized search comprising the steps of:

determining a query;

determining a user;

determining a user history of access patterns within an information repository of linked information elements;

determining a user profile based on the user history;

determining a proximal neighborhood based on the user history; and

determining personalized search results based on the query, the user profile and the proximal neighborhood.

2. The method of claim 1, in which the proximal neighborhood is determined based on crawling parameters.

3. The method of claim 2, in which the crawling parameters specifies a crawling strategy from at least one of:

the PageRank function;

the backlink function; and

breadth first search.

4. The method of claim 2, in which at least one of the crawling parameters is a threshold link distance.

5. The method of claim 1, in which the user history is from at least one of:

a web access log;

a library manager;

a web proxy server log;

a browser history file; and

a server access log.

6. The method of claim 1, in which the information repository is at least one of:

the world wide web, a document repository and a hard drive.

7. The method of claim 6, in which the document repository is at least one of:

a digital library, document management system, and a server.

8. The method of claim 1, in which user histories, user profiles and proximal neighborhoods are determined for a first user and at least one other user, and in which at least one of the user history, the user profiles and the proximal neighborhoods are combined.

9. The method of claim 8, in which the combination is at least one of:

union, intersection, addition and subtraction.

10. The method of claim 1, in which a first user history is associated with a first user and a second user history is associated with a second user and the first user history is used by the second user.

11. The method of claim 10, in which a first user proximal neighborhood is associated with a first user and a second user proximal neighborhood is associated with a second user and the first user proximal neighborhood is used by the second user.

12. The method of claim 10, in which a first user profile is associated with a first user and a second user profile is associated with a second user and the first user profile is used by the second user.

13. The method of claim 1, in which the linked information elements are comprised of at least one of:

audio, video and textual information;

14. The method of claim 1, in which the linked information elements are linked documents.

15. The method of claim 1, are comprised of at least one of:

audio, video and textual information.

16. The method of claim 1, in which the search results are ordered based on the user profile.

17. The method of claim 1, in which the user history is categorized.

18. The method of claim 17, in which the categories are based on at least one of:

tasks, projects, organizations.

19. A system for personalized search comprising:

a memory;

an input/output circuit that receives a user query;

a processor that determines a relevant user;

a user history determination circuit that determines the user history within a repository of linked information elements;

a profile determination circuit that determines a user profile descriptive of the linked information elements in the user history;

a proximal neighborhood determination circuit that determines a proximal neighborhood of unseen information elements from the repository linked to user history information element within at least a threshold distance metric;

a search circuit that searches the information elements in the proximal neighborhood for the query; and

a ranking circuit that ranks the search results based on the user profile.

20. The system of claim 19, in which the proximal neighborhood is determined based on crawling parameters.

21. The system of claim 20, in which the crawling parameters specifies a crawling strategy from at least one of:

the PageRank function;

the backlink function; and

breadth first search.

22. The system of claim 20, in which at least one of the crawling parameters is a threshold link distance.

23. The system of claim 19, in which the user history is from at least one of:

a web access log;

a library manager;

a web proxy server log;

a browser history file; and

a server access log.

24. The system of claim 19, in which the information repository is at least one of:

the world wide web, a document repository and a hard drive.

25. The system of claim 24, in which the document repository is at least one of:

a digital library, document management system, and a server.

26. The system of claim 19, in which user histories, user profiles and proximal neighborhoods are determined for a first user and at least one other user, and in which at least one of the user history, the user profiles and the proximal neighborhoods are combined.

27. The system of claim 26, in which the combination is at least one of:

union, intersection, addition and subtraction.

28. The system of claim 19, in which a first user history is associated with a first user and a second user history is associated with a second user and the first user history is used by the second user.

29. The system of claim 28, in which a first user proximal neighborhood is associated with a first user and a second user proximal neighborhood is associated with a second user and the first user proximal neighborhood is used by the second user.

30. The system of claim 28, in which a first user profile is associated with a first user, and a second user profile is associated with a second user, and the first user profile is used by the second user.

31. The system of claim 19, in which the linked information elements are comprised of at least one of:

audio, video and textual information;

32. The system of claim 19, in which the linked information elements are linked documents.

33. The system of claim 19 are comprised of at least one of:

audio, video and textual information.

34. The system of claim 19, in which the search results are ordered based on the user profile.

35. The system of claim 19, in which the user history is categorized.

36. The system of claim 35, in which the categories are based on at least one of:

tasks, projects, organizations.

37. A carrier wave encoded to transmit a control program, useable to program a computer to provide personalized search, to a device for executing the program, the control program comprising:

instructions for determining a query;

instructions for determining a user;

instructions for determining a user history of access patterns within an information repository of linked information elements;

instructions for determining a user profile based on the user history;

instructions for determining a proximal neighborhood based on the user history; and

instructions for determining personalized search results based on the query, the user profile and the proximal neighborhood.

38. A computer readable storage medium comprising computer readable program code embodied on the computer readable storage medium, the computer readable program code useable to program a computer to provide personalized search comprising the steps of:

determining a query;

determining a user;

determining a user history of access patterns within an information repository of linked information elements;

determining a user profile based on the user history;

determining a proximal neighborhood based on the user history; and

determining personalized search results based on the query, the user profile and the proximal neighborhood.