Dynamic Search with Implicit User Intention Mining

• US 20070156647A1
• Filed: 12/29/2005
• Published: 07/05/2007
• Est. Priority Date: 12/29/2005
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: Li, Shipeng, Shen, Guo
• Granted Patent: US 7,599,918 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: Y10S 707/99933 Query processing, i.e. searching
• CPC Class Codes: Y10S 707/99936 Pattern matching access
• CPC Class Codes: Y10S 707/99942 Manipulating data structure, e.g. compression, compaction, compilation

Abstract

After a user instigated search returns results, an intention mining engine collects information from the natural user responses to the results. This information is used to refine the search.

Claims

1. A computer comprising:

a processor;

a memory coupled to the processor, wherein the memory includes computer-program instructions executable by the processor, the computer-program instructions comprising a user intention mining engine to collect information from natural user responses to search results returned by a search instigated by the user and utilize the information to refine the search.

2. The computer of claim 1, wherein the user intention mining module further collects information from natural user responses comprising information regarding the interests, preferences and behavior of the user available outside of the search.

3. The computer of claim 2, wherein the information from natural user responses comprises metadata or other information from websites visited by the user.

4. The computer of claim 1, wherein the information from natural user responses comprises one of a number of user clicks a user makes on a particular result returned by the search, which results returned by the search are selected by the user, and the length of time a user spends reviewing a particular returned result.

5. The computer of claim 1, wherein the user intention mining module further collects information from natural user responses comprising the length of a search term input by the user.

6. The computer of claim 1, wherein the user intention mining module further collects information from natural user responses comprising initial and successive search terms input by the user.

7. The computer of claim 1, wherein the user intention mining module further collects information from natural responses including one of information from files access by the user, information from emails and instant messages sent or received by the user at the computer, and information input to the computer by the user.

8. One or more computer-readable media comprising computer-program instructions executable by a processor, the computer-program instructions including instructions for:

obtaining a search term from a user;

performing a search based on the search term;

presenting search results based on the search term in a primary window;

mining natural responses of a user to the search terms;

refining the search based on the natural responses;

returning refined search results; and

displaying the results of the refined search to the user.

9. The one or more computer-readable media of claim 8, further comprising collecting information regarding user interests, preferences and behavior available outside of the search.

10. The one or more computer-readable media of claim 8, further comprising collecting information regarding user preferences from both a server and a client device on which the user has been active.

11. The one or more computer-readable media of claim 8, wherein refining comprises refining the search after each natural response is made by the user.

12. The one or more computer-readable media of claim 8, wherein refining comprises one of refining the search after a preset number of natural responses are made by the user, and refining the search after a preset period of time has elapsed.

13. The one or more computer-readable media of claim 8, wherein returning comprises presenting the refined search results in an updating window.

14. The one or more computer-readable media of claim 8, further comprising collecting information from other applications with which the user interacts during the search.

19. The method of claim 14, wherein returning comprises returning results of a refined search based on the natural responses in a second window.

20. The method of claim 19, further comprising:

allowing the user to interact with results in the second window;

mining interest and behavior information concerning the user from user interactions with the search results in the second window; and

presenting results from a further refined search in the first window.

15. A computer-based method, comprising:

displaying results of a search instigated by the user through input of one or more search terms in a first window;

enabling the user to interact with the search results;

mining interest and behavior information concerning the user from interactions of the user with the search results; and

returning results of a refined search based on the natural responses.

16. The method of claim 15, wherein mining comprises tracking a number of clicks made by the user on a particular search result.

17. The method of claim 15, wherein mining comprises monitoring the amount of time elapsing between a click on a search result, and a click on another search result.

18. The method of claim 15, further comprising reviewing a user'"'"'s files and a user'"'"'s actions outside of the search to develop user interest and behavior information.