METHODS AND APPARATUS FOR CONTEXT-SENSITIVE INFORMATION RETRIEVAL BASED ON INTERACTIVE USER NOTES

• US 20100106485A1
• Filed: 10/24/2008
• Published: 04/29/2010
• Est. Priority Date: 10/24/2008
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Zhou, Michelle X., Lu, Jie
• Granted Patent: US 8,671,096 B2
• Time in Patent Office: Days
• US Class Current: 704/9
• CPC Class Codes: G06F 16/243 Natural language query formulation
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising

Abstract

Information retrieval systems and methods are provided based on interactive user notes. Information is retrieved from one or more data sources based on user notes by obtaining the user notes containing one or more information requests; identifying the one or more information requests from the user notes; interpreting at least one of the information requests in context; generating one or more queries required for the at least one interpreted information request; identifying an update to the user notes, the update containing one or more updated information requests; and processing the updated user notes to generate one or more queries required for the updated information requests. If the user notes contain multiple information requests, at least one query is generated for each of the plurality of information requests. The information requests can be interpreted based on user-specified context guides.

Claims

1. A method for retrieving information from one or more data sources based on user notes, comprising:

obtaining said user notes containing one or more information requests;

identifying said one or more information requests from said user notes;

interpreting at least one of said information requests in context;

generating one or more queries required for said at least one interpreted information request;

identifying an update to said user notes, said update containing one or more updated information requests; and

processing said updated user notes to generate one or more queries required for said updated information requests.

2. The method of claim 1, wherein said step of identifying said one or more information requests further comprises the step of identifying multiple information requests from said user notes using one or more of syntactic cues, a structure of said user notes, a format of said user notes, and semantic rules.

3. The method of claim 1, further comprising the step of interpreting said information requests in context to generate a semantic representation that encodes a meaning of said information requests.

4. The method of claim 3, further comprising the step of identifying a set of semantic elements from said information requests including one or more data concepts, data attributes and data constraints.

5. The method of claim 4, further comprising the step of inferring a semantic relationship between said set of semantic elements.

6. The method of claim 4, further comprising the step of generating a semantic representation of said information request to encode said semantic elements and semantic relationships.

7. The method of claim 6, wherein said semantic representation is a graph-based model.

8. The method of claim 7, wherein said graph-based model comprises an anchor node defining a type of information to be retrieved.

9. The method of claim 7, wherein each node in said graph-based model represents a data concept and contains one or more data attributes and one or more data constraints.

10. The method of claim 7, wherein each link in said graph-based model represents a semantic relationship between two of said data concepts.

11. The method of claim 3, further comprising the step of integrating a context of existing user notes with said semantic representation of a current information request.

12. The method of claim 11, wherein said context integration further comprises the steps of identifying context integration candidates from one or more previously derived information requests based on one or more of a semantics of said information requests and a structure of said user notes, and determining integration operations to merge said current information request with said context integration candidates.

13. The method of claim 1, wherein said generation of one or more queries further comprises the step of determining query types and one or more query parameters for said one or more queries that satisfy said information request.

14. The method of claim 13, further comprising the step of using a procedural method to formulate a query of a specific type for said information request using said one or more query parameters.

15. A method for retrieving information from one or more data sources based on user notes, comprising:

obtaining said user notes containing a plurality of information requests;

identifying said plurality of information requests from said user notes;

interpreting said plurality of said information requests in context; and

generating one or more queries for each of said plurality of interpreted information requests.

16. The method of claim 15, wherein said step of identifying said plurality of information requests further comprises the step of identifying multiple information requests from said user notes using one or more of syntactic cues, a structure of said user notes, a format of said user notes, and semantic rules.

17. The method of claim 15, further comprising the step of interpreting said information requests in context to generate a semantic representation that encodes a meaning of said information requests.

18. A method for retrieving information from one or more data sources based on user notes, comprising:

obtaining said user notes containing one or more information requests and one or more user-specified context guides;

identifying said one or more information requests from said user notes;

interpreting at least one of said information requests based on said one or more user-specified context guides; and

generating one or more queries required for said at least one interpreted information request.

19. The method of claim 18, wherein said one or more user-specified context guides comprises one or more of syntactic cues, a structure of said user notes, and a format of said user notes.

20. The method of claim 18, further comprising the step of inferring at least a portion of said user notes based on a prior user note.