TWO STAGE SEARCH

• US 20120109949A1
• Filed: 01/04/2012
• Published: 05/03/2012
• Est. Priority Date: 11/14/2005
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: CAO, Yunbo, LI, Hang
• Granted Patent: US 8,849,787 B2
• Time in Patent Office: Days
• US Class Current: 707/728
• CPC Class Codes: G06F 16/3344 using natural language analysis

Abstract

A two stage model identifies individuals having knowledge in a subject matter area relevant to a query. A relevance model receives a query and identifies documents, or other information, relevant to the query. A co-occurrence model identifies individuals, in the retrieved documents, related to the subject matter of the query. Individuals identified can be scored by combining scores from the relevance model and the co-occurrence model and output in a rank ordered list.

Claims

1. A search system, comprising:

a relevance model stored on a computer readable medium;

a computer processor that applies the relevance model to a query so as to identify information relevant to the query; and

a co-occurrence model stored on the computer readable medium, wherein the computer processor applies the co-occurrence model to the identified information so as to identify a plurality of persons related to the query based on the query and the identified information.

2. The system of claim 1, further comprising a ranking component that ranks the plurality of persons based on said application of the co-occurrence model.

3. The system of claim 2, further comprising a re-ranking component that computes a score for each of the plurality of persons based on a relationship with another of the plurality of persons.

4. The system of claim 3, wherein the re-ranking component re-ranks the plurality of persons based on said computed scores.

5. The system of claim 1, wherein the co-occurrence model identifies co-occurrences of the subject matter of the query and a candidate person, the identified plurality of persons including the candidate person.

6. The system of claim 1, wherein the information relevant to the query comprises documents and related metadata, and wherein the computer processor applies the co-occurrence model to the identified information with programmatic consideration given to the related metadata.

7. The system of claim 6, wherein the related metadata is document body metadata.

8. The system of claim 6, wherein the related metadata is title metadata.

9. The system of claim 6, wherein the related metadata is author metadata.

10. The system of claim 6, wherein the related metadata is anchor text metadata.

11. The system of claim 6, wherein the related metadata is indicative of a hierarchy of document sections.

12. A computer implemented method of identifying persons related to a query, comprising:

processing a query to obtain a query result;

identifying a set of people based on a relationship of the set of people to the query, the relationship being determined based at least in part on information derived from the query result; and

outputting the set of people.

13. The method of claim 12, further comprising ranking the set of people such that outputting the set of people comprises outputting the set of people in a ranked order.

14. The method of claim 13, wherein ranking comprises ranking based at least in part on a score computed based on co-occurrences of information in the query result and people included in the set of people.

15. The method of claim 13, wherein ranking comprises ranking at least in part based on co-occurrences of information in the query result and people included in the set of people.

16. The method of claim 13, wherein ranking comprises ranking based at least in part on a proximity a person in the set of people relative to a portion of information in the query result.

17. A computer implemented method of identifying persons related to a query, comprising:

processing a query to obtain a query result;

identifying a set of people based on a number of co-occurrences of a person with information derived from the query result; and

outputting the set of people ranked based at least in part on the number of co-occurrences.

18. The method of claim 17, wherein the query result includes documents and related metadata, and wherein the identification of the set of people is based at least in part on the related metadata.

19. The method of claim 18, wherein the metadata is author metadata.

20. The method of claim 18, wherein the metadata anchor text metadata.