System and method for ranking documents through human assistance

• US 8,352,464 B2
• Filed: 06/09/2010
• Issued: 01/08/2013
• Est. Priority Date: 07/24/2009
• Status: Active Grant

• Current Assignee: Peer Belt Inc.
• Original Assignee: Peer Belt Inc.
• Inventors: Fotev, Krassimir
• Primary Examiner(s): Wong, Leslie
• Application Number:
  US12/797,376
• Publication Number:
  US 20110022578A1
• Time in Patent Office: 944 Days
• Field of Search: 707723-728, 707/999.003, 707/999.01, 707/999.107, 707/706, 707/707
• US Class Current: 707/723
• CPC Class Codes: G06F 16/3326 using relevance feedback from the user, e.g. relevance feedback on documents, documents sets, document terms or passages
• CPC Class Codes: G06F 16/3331 Query processing
• CPC Class Codes: G06F 16/338 Presentation of query results

Abstract

Embodiments of system and method are provided for ranking contents of one or more documents through human assistance. The system includes one or more search engines configured to provide one or more documents based on a search query sent by a user. Further, the system includes a client application configured to capture the query sent by the user and the document tags provided by the document author or inbound links. The client application monitors one or more interactions of the user with the one or more documents for one or more events and provides a relevance measure to the one or more documents based on the captured query or tags, and the one or more events.

Claims

1. A system for ranking contents of one or more documents, the system comprising:

one or more search engines configured to provide one or more documents based on a search query sent by a user; and

one or more processors executing a client application and configured to;

capture the query sent by the user;

capture one or more document tags;

determine, with respect to at least one of the one or more documents, that the user has navigated the at least one of the one or more documents for at least a defined period of time, the defined period of time being defined based on at least one of (a) a length of the at least one of the one or more documents, (b) a word count of the at least one of the one or more documents, and (c) a fragmentation of the at least one of the one or more documents;

monitor, based on a determination that the user has navigated the at least one of the one or more documents for at least the defined period of time, in relation to the at least one of the one or more documents, and throughout one or more time intervals, one or more interaction events, the one or more interaction events comprising one or more behaviors that demonstrate interest in the at least one of the one or more documents, each of the one or more interaction events having a respective weight that reflects a relative significance of the interaction event with respect to a relevance of the at least one of the one or more documents to the query; and

compute, in relation to at least one of the query and the one or more tags, a relevance measure for the at least one of the one or more documents based on the one or more interaction events, the respective weights of each of the one or more interaction events, and the one or more time intervals.

2. The system of claim 1 further comprising:

one or more promotion stores configured to;

store the at least one of the one or more ranked documents;

store one or more vectors, wherein the vectors are based on at least one of (a) the captured query and the at least one of the one or more documents and (b) the one or more document tags and the at least one of the one or more documents; and

store one or more promotion counters associated with the one or more vectors.

3. The system of claim 1, wherein the client application comprises a browser plug-in.

4. The system of claim 1, wherein the client application comprises a browser.

5. The system of claim 1, wherein the client application is further configured to:

determine whether the relevance measure is above a predetermined critical value;

link the at least one of the one or more documents to a perceived semantic domain when the relevance measure is greater than the predetermined critical value; and

persist the semantic links to the at least one of the one or more documents.

6. The system of claim 5, wherein the semantic domain is perceived based on the captured query.

7. The system of claim 1, wherein the one or more behaviors comprise one or more of:

a document printing, a document saving, a document copying, a document bookmarking, a document address copying, a number of document visits, a keyboard data entry, a document scrolling, a pointer clicking, a document zooming, a document switching, time spent in reading the document, and a webpage navigation.

8. The system of claim 1, wherein the one or more time intervals are normalized based on a length of the at least one of the one or more documents.

9. The system of claim 1, further comprising a display configured to display one or more search results to the user based on the relevance measure.

10. A computer-implemented method for ranking contents of one or more documents, the method comprising:

capturing, with one or more processors executing a client application, a query sent by a user to a search engine;

capturing one or more tags provided for the one or more documents;

determining, with respect to at least one of the one or more documents, that the user has navigated the at least one of the one or more documents for at least a defined period of time, the defined period of time being defined based on at least one of (a) a length of the at least one of the one or more documents, (b) a word count of the at least one of the one or more documents, and (c) a fragmentation of the at least one of the one or more documents;

monitoring, based on a determination that the user has navigated the at least one of the one or more documents for at least the defined period of time, in relation to the at least one of the one or more documents and throughout one or more time intervals, one or more interaction events, the one or more interaction events comprising one or more behaviors that demonstrate interest in the at least one of the one or more documents, each of the one or more interaction events having a respective weight that reflects a relative significance of the interaction event with respect to a relevance of the at least one of the one or more documents to the query; and

computing, in relation to at least one of the query and the one or more tags, a relevance measure for the at least one of the one or more documents based on the one or more interaction events, the respective weights of each of the one or more interaction events, and the one or more time intervals.

11. The method of claim 10 further comprising sending at least one of (a) the captured query and (b) the tags to one or more promotion stores for the at least one of the one or more documents.

12. The method of claim 10, wherein the client application comprises a browser plug-in.

13. The method of claim 10, wherein the client application comprises a browser.

14. The method of claim 10 further comprising determining whether the relevance measure is above a predetermined critical value.

15. The method of claim 14 further comprising linking the at least one of the one or more documents to a perceived semantic domain when the relevance measure is greater than the predetermined critical value.

16. The method of claim 15, wherein the semantic domain is perceived based on the captured query.

17. The method of claim 15 further comprising persisting a semantic link to the at least one of the one or more documents.

18. The method of claim 10, wherein the one or more behaviors comprise one or more of:

a document printing, a document saving, a document bookmarking, a document address copying, a number of document visits, a document scrolling, a pointer clicking, a document zooming, a document switching, and a time spent reading the document.

19. The method of claim 10, wherein the one or more time intervals are normalized based on a length of the at least one of the one or more documents.

20. The method of claim 10, further comprising supplementing one or more search results provided to the user based on the relevance measure.

21. A non-transitory computer-readable medium containing instructions executable by a processor for performing a method for ranking contents of one or more documents, the method comprising:

capturing a query sent by a user to a search engine;

capturing one or more tags provided for the one or more documents;

determining, with respect to at least one of the one or more documents, that the user has navigated the at least one of the one or more documents for at least a defined period of time, the defined period of time being defined based on at least one of (a) a length of the at least one of the one or more documents, (b) a word count of the at least one of the one or more documents, and (c) a fragmentation of the at least one of the one or more documents;

monitoring, based on a determination that the user has navigated the at least one of the one or more documents, for at least the defined period of time, in relation to the at least one of the one or more documents and throughout one or more time intervals, one or more interaction events, the one or more interaction events comprising one or more behaviors that demonstrate interest in the at least one of the one or more documents each of the one or more interaction events having a respective weight that reflects a relative significance of the interaction event with respect to a relevance of the at least one of the one or more documents to the query; and

computing, in relation to at least one of the query and the one or more tags, a relevance measure for the at least one of the one or more documents based on the one or more interaction events, the respective weights of each of the one or more interaction events, and the one or more time intervals.

22. The computer-readable medium of claim 21, wherein the method further comprises:

determining whether the relevance measure is above a predetermined critical value;

based on the determining, linking the at least one of the one or more documents to a perceived semantic domain; and

persisting the semantic link to the at least one of the one or more documents.