Adaptive dissemination of personalized and contextually relevant information

• US 7,577,718 B2
• Filed: 07/31/2006
• Issued: 08/18/2009
• Est. Priority Date: 07/31/2006
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: Chandrasekar, Raman, Forney, Michael K., Slawson, Dean A.
• Primary Examiner(s): Meky; Moustafa M.
• Assistant Examiner(s): Williams; Clayton R
• Application Number:
  US11/461,388
• Publication Number:
  US 20080028036A1
• Time in Patent Office: 1,114 Days
• Field of Search: None
• US Class Current: 709/218
• CPC Class Codes: G06F 16/9535 Search customisation based on user profiles and personalisation
• CPC Class Codes: H04L 67/02 based on web technology, e.g. hypertext transfer protocol [HTTP]
• CPC Class Codes: Y10S 707/99933 Query processing, i.e. searching
• CPC Class Codes: Y10S 707/99937 Sorting

Abstract

An information dissemination system identifies the interests of a user from information derived from various sources, such as, by way of example, the user'"'"'s calendar entries, emails, World Wide Web (web) usage information, documents, enterprise resource planning (ERP) data, and the like. The information dissemination system then creates a profile for the user, and includes in the profile the details about the user'"'"'s interests including the information sources to query, the user events which were analyzed to determine the user'"'"'s interest, the topics derived from the analyzed events, the queries generated for the analyzed events, a language model of terms that are relevant to the user, and other information regarding the user. The information dissemination system then periodically executes the queries against the appropriate information sources, ranks the search results obtained from executing the queries, and renders the ranked search results to the user.

Claims

1. A computer-implemented method performed by a computer having a memory and a processor for disseminating contextually relevant information based on user interests, the method comprising:

identifying user interests of a user by extracting terms from files associated with the user;

automatically generating a user profile based on the user interests by,

generating clusters of similar terms from the extracted terms,

associating at least one cluster with each of a plurality of events associated with the user, and

for each event associated with the user,

for each identified cluster associated with the event,

identifying key terms in the cluster,

generating a query from the key terms,

generating a tome that includes the generated query and event information associated with the event, and

placing the generated topic in the user profile so that the user profile comprises key terms derived from terms extracted from the user interests and at least one query generated from one or more key terms; and

periodically generating information for rendering to the user by,

for each topic in the user profile, obtaining search results by executing each query associated with the topic against information sources,

for each search result,

identifying an event associated with the search result,

calculating a static weight for the search result,

calculating a temporal weight for the search result based at least in part on the proximity of the identified event associated with the search result to the current time,

calculating an adaptive weight for the search result based at least in part on a user preference of the user,

calculating a total weight for the search result based at least in part on the static, temporal, and adaptive weights, and

ranking the search results based on the total weight calculated for each search result, and

rendering the ranked search results to the user

wherein the generating and garnering are performed by the processor executing instructions stored in the memory.

2. The method of claim 1, wherein at least one user interest is identified from an appointment of the user.

3. The method of claim 1, wherein at least one user interest is identified from an email of the user.

4. The method of claim 1, wherein the key terms are derived based on overall weights corresponding to the terms.

5. The method of claim 4, wherein the overall weight corresponding to a term is based in part on a frequency of the term.

6. The method of claim 4, wherein the overall weight corresponding to a term is based in part on a weight of the term.

7. The method of claim 4, wherein the overall weight corresponding to a term is based in part on a relevance of the term to the user interests.

8. The method of claim 1 further comprising:

receiving feedback from the user regarding the rendered search results; and

incorporating the feedback into the user profile as user interests.

9. A computer-readable medium containing computer-executable instructions that when executed by a processor of a computer system cause the computer system to perform a method for adaptively disseminating contextually relevant information based on user interests of a user, the method comprising:

identifying user interests from clusters of terms extracted from one or more user emails, each cluster of terms corresponding to a set of similar terms;

identifying a user event;

identifying topics based on the user event;

for each topic,

identifying clusters of terms for the topic;

for each of the identified clusters of terms, identifying key terms in the cluster by,

selecting a term in the cluster,

calculating a type weight for the selected term,

calculating a relevance weight for the selected term,

calculating an overall weight for the selected term based at least in part on the calculated type and relevance weights, and

if the overall weight exceeds a predetermined threshold, identifying the selected term as a key term; and

generating a query based on the key terms;

obtaining search results by executing the queries associated with each topic

against information sources;

ranking the search results; and

rendering the ranked search results to the user.

10. The computer-readable medium of claim 9, wherein the overall weight of a term is based in part on a frequency of the term in the clusters of terms.

11. The computer-readable medium of claim 9, wherein the overall weight of a term is based in part on a weight of the term in the clusters of terms.

12. The computer-readable medium of claim 9, wherein the relevance weight of a term is based in part on a relevance of the term in the clusters of terms to the user interests.

13. The computer-readable medium of claim 9 further comprising:

receiving feedback from the user regarding the rendered search results; and

incorporating the feedback into the user interests.

14. The computer-readable medium of claim 9, wherein the query is based at most on the six highest ranked key terms according to overall weight.

15. The computer-readable medium of claim 9, wherein the search results are ranked based in part on static weights corresponding to the search results.

16. The computer-readable medium of claim 9, wherein the search results are ranked based in part on adaptive weights corresponding to the search results.

17. The computer-readable medium of claim 9, wherein ranking the search results includes:

calculating a static weight for each search result;

calculating a temporal weight for each search result;

calculating an adaptive weight for each search result;

calculating a total weight for each search result based at least in part on the static, temporal, and adaptive weights; and

ranking the search results based on the total weights of each search result.

18. An information dissemination system comprising:

a profiler component that executes on a client and that identifies user interests, that derives key terms from terms extracted from the user interests, and that generates queries from one or more of the key terms, wherein the key terms are derived based on the relevance of the terms to the user interests;

an information garnerer component that executes on a server and that obtains search results from executing the queries against information sources;

a ranker component that executes on the server and that ranks the search results by,

calculating a static weight for each search result,

calculating a temporal weight for each search result,

calculating a total weight for each search result based at least in part on the static and temporal weights, and

ranking the search results based on the total weight of each search result; and

a renderer component that executes on the server and that renders the ranked search results for viewing.

19. The system of claim 18 further comprising a feedback component that executes on the server and that incorporates feedback regarding the search results into the user interests.

20. The system of claim 18, wherein calculating a temporal weight for a search result includes:

identifying an event associated with the search result;

determining a current time;

determining a start time for the event;

determining an end time for the event;

when the current time is earlier than the determined end time, calculating a temporal weight based at least in part on the difference between the current time and the start time; and

when the current time is later than the determined end time, calculating a temporal weight based at least in part on the difference between the current time and the determined end time.