Method and system for adapting search results to personal information needs

• US 20060259480A1
• Filed: 05/10/2005
• Published: 11/16/2006
• Est. Priority Date: 05/10/2005
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: Ma, Wei-Ying, Chen, Zheng, Zhang, Benyu, Zeng, Hua-Jun, Jiang, Xue-Mei, Xue, Gui-Rong
• Granted Patent: US 7,630,976 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 16/9558 Details of hyperlinks; Management of linked annotations
• CPC Class Codes: Y10S 707/99935 Query augmenting and refining, e.g. inexact access

Abstract

A method and system for adapting search results of a query to the information needs of the user submitting the query is provided. A search system analyzes click-through triplets indicating that a user submitted a query and that the user selected a document from the results of the query. To overcome the large size and sparseness of the click-through data, the search system when presented with an input triplet comprising a user, a query, and a document determines a probability that the user will find the input document important by smoothing the click-through triplets. The search system then orders documents of the result based on the probability of their importance to the input user.

Claims

1. A method in a computer system for calculating importance of a document, the method comprising:

providing user, query, and document triplets indicating that the user submitted the query and that the user selected the document from a result of the query;

receiving a user, query, and document as an input triplet; and

determining a probability that the user will find the input document important by performing a smoothing of the provided triplets to account for triplets not provided and calculating the probability based on the smoothing.

2. The method of claim 1 wherein the smoothing uses a back-off technique.

3. The method of claim 1 wherein the probability is the probability of the input document given the input user and input query when the provided triplets contain the input user, input query, and input document in a triplet and is the probability of the input document given the input query when the provided triplets do not contain the input user, input query, and input document in a triplet.

4. The method of claim 1 wherein the probability is based on the probability that the input document is in a document cluster and the probability of the document cluster given a user cluster and query cluster.

5. The method of claim 1 wherein the probability is based on probabilities of clusters identified based on relationships between users, queries, and documents.

6. The method of claim 1 wherein the probability is based on similarity between an input query and query clusters identified based on relationships between users and documents.

7. The method of claim 1 wherein the probability is based on similarity between an input document and document clusters identified based on relationships between users and queries.

8. The method of claim 1 wherein the smoothing uses a combination of back-off, clustering, and content similarity techniques.

9. A computer-readable medium containing instructions for controlling a computer system to calculate relevance of a document to a user, by a method comprising:

providing click-through data;

identifying user, query, and document triplets from the click-through data indicating that the user submitted the query and the user selected the document from results of the query;

receiving a user, query, and document as an input triplet; and

determining a probability that the input user will find the input document relevant by performing a smoothing of the identified triplets to account for sparseness of the triplets and calculating the probability based on the smoothed triplets.

10. The computer-readable medium of claim 9 including processing the triplets to identify clusters of users, queries, and documents.

11. The computer-readable medium of claim 9 including processing the triplets to calculate various counts.

12. The computer-readable medium of claim 9 wherein the probability is the probability of the input document given the input user and input query when the user, query, and document are in an identified triplet and is the probability of the input document given the input query otherwise.

13. The computer-readable medium of claim 9 wherein the probability is based on the probability of the input document given a document cluster and the probability of the document cluster given a user cluster and query cluster.

14. The computer-readable medium of claim 9 wherein the probability is based on similarity between the input document and document clusters identified based on relationships between users and queries.

15. The computer-readable medium of claim 9 wherein the smoothing uses a combination of back-off, clustering, and content similarity techniques.

16. A computer system for calculating relevance of a document, comprising:

a click-through data store;

a component that identifies user, query, and document triplets from the click-through data;

a component that receives a user, query, and documents as input triplets; and

determining a probability that the input user will find each document relevant by performing a smoothing of the received triplets to account for sparseness of the triplets and calculating the probability based on the smoothed triplets.

17. The computer system of claim 16 including a component to identify clusters of users, queries, and documents based on relationships between users, queries, and documents.

18. The computer system of claim 16 wherein the probability is the probability of the input document given the input user and input query when the user, query, and document triplet is received and is the probability of the input document given the input query otherwise and wherein a back-off factor is applied to the probability.

19. The computer system of claim 16 wherein the probability is based on the probability of the input document given a document cluster and the probability of the document cluster given a user cluster and a query cluster.

20. The computer system of claim 16 wherein the probability is based on similarity between an input document and document clusters identified based on relationships between users and queries.