Method and apparatus for identifying and extracting meaning in documents

• US 8,583,580 B2
• Filed: 01/24/2011
• Issued: 11/12/2013
• Est. Priority Date: 10/10/2007
• Status: Active Grant

• Current Assignee: Northern Light Group, LLC
• Original Assignee: Northern Light Group, LLC
• Inventors: Seuss, C. David
• Primary Examiner(s): Gaffin, Jeffrey A
• Assistant Examiner(s): CHANG, LI WU
• Application Number:
  US13/012,395
• Publication Number:
  US 20110119271A1
• Time in Patent Office: 1,023 Days
• Field of Search: None
• US Class Current: 706/47
• CPC Class Codes: G06N 5/022 Knowledge engineering; Knowledge acquisition

Abstract

Systems and methods are provided for an expert search engine. In one embodiment, a method may employ a computer based system to define a meaning taxonomy including one or more meaning loaded entities, associate documents to the meaning taxonomy based on a set of expert rules and the syntactic structures included in the documents and provide the results to a user. In another embodiment, a system includes interfaces for meaning taxonomy, expert rule and search criteria information. The information provided by these interfaces is then used to determine how a set of documents maps to meaning taxonomy to assist the user in locating documents that are most interesting from a meaning perspective and to provide a user with an interface to navigate the meaning taxonomy as it applies to the set of documents.

Claims

1. A computer-implemented method for extracting meaning from a plurality of documents, the method comprising:

identifying, by a computer, a meaning taxonomy including a plurality of selected concepts;

identifying a group of syntactic structures including at least one syntactic structure;

associating at least one of the group of syntactic structures with at least one selected concept of the plurality of selected concepts;

applying at least one expert rule selected from a group of expert rules to at least one document of the plurality of documents, the group of expert rules being associated with the at least one selected concept, the at least one expert rule including a plurality of logical propositions, at least one logical proposition of the plurality of logical propositions including an evaluation of whether an association exists between one or more of the group of syntactic structures associated with the at least one selected concept and one or more syntactic structures included in the at least one document of the plurality of documents; and

associating, responsive to existence of the association, the at least one document of the plurality of documents with the at least one selected concept of the plurality of selected concepts.

2. The method according to claim 1, wherein identifying the meaning taxonomy comprises identifying a plurality of meaning loaded entities.

3. The method according to claim 1, wherein identifying a group of syntactic structures comprises identifying a word.

4. The method according to claim 1, wherein identifying a group of syntactic structures comprises identifying a phrase.

5. The method according to claim 1, wherein identifying a group of syntactic structures comprises identifying a sentence.

6. The method according to claim 1, wherein associating the at least one of the group of syntactic structures comprises associating a plurality of the group of syntactic structures with the at least one selected concept of the plurality of selected concepts.

7. The method according to claim 1, wherein associating the at least one of the group of syntactic structures comprises associating a plurality of the group of syntactic structures in relative proximity to each other with the at least one selected concept of the plurality of selected concepts.

8. The method according to claim 1, wherein associating the at least one document of the plurality of documents comprises associating the at least one document of the plurality of documents with the at least one selected concept of the plurality of selected concepts, the at least one document of the plurality of documents including at least one syntactic structure of the group of syntactic structures that is associated with the at least one selected concept of the plurality of selected concepts.

9. The method according to claim 1, wherein associating the at least one document of the plurality of documents comprises associating the at least one document of the plurality of documents with the at least one selected concept of the plurality of selected concepts, the at least one document of the plurality of documents including none of the group of syntactic structures that is associated with the at least one selected concept of the plurality of selected concepts.

10. The method according to claim 1, wherein associating the at least one of the group of syntactic structures with the at least one selected concept of the plurality of selected concepts comprises:

associating at least one normalized syntactic structure with the at least one selected concept of the plurality of selected concepts; and

associating at least one non-normalized syntactic structure with the at least one normalized syntactic structure;

wherein associating the at least one document of the plurality of documents comprises associating the at least one document of the plurality of documents with at least one selected concept of the plurality of selected concepts based on an association between the at least one document of the plurality of documents and at least one non-normalized syntactic structure of the group of syntactic structures.

11. The computer-implemented method according to claim 1, further comprising:

executing a search against the plurality of documents, the search resulting in a subset of the plurality of documents; and

providing the subset of the plurality of documents to an external entity, the subset of the plurality of documents grouped by at least one selected concept of the plurality of selected concepts that is associated with at least one document of the subset of the plurality of documents.

12. A non-transitory computer readable medium having computer readable signals stored thereon that define instructions that, as a result of being executed by a computer, instruct the computer to perform a method for extracting meaning from a plurality of documents, the method comprising:

identifying a meaning taxonomy including a plurality of selected concepts;

identifying a group of syntactic structures including at least one syntactic structure;

associating at least one of the group of syntactic structures with at least one selected concept of the plurality of selected concepts;

applying at least one expert rule selected from a group of expert rules to at least one of the plurality of documents, the group of expert rules being associated with the at least one selected concept, the at least one expert rule including a plurality of logical propositions, at least one logical proposition of the plurality of logical propositions including an evaluation of whether an association exists between one or more of the group of syntactic structures associated with the at least one selected concept and one or more syntactic structures included in the at least one document of the plurality of documents; and

associating, responsive to existence of the association, at least one of the plurality of documents with the at least one selected concept of the plurality of selected concepts.

13. A computer-implemented method for configuring a meaning taxonomy comprising at least one meaning loaded entity, the method comprising:

providing, by a computer, an input component that allows a user to provide at least one non-normalized syntactic structure;

receiving the at least one non-normalized syntactic structure;

providing an input component that allows the user to build provide at least one expert rule, the at least one expert rule including a plurality of logical propositions;

receiving the at least one expert rule;

providing an input component that allows the user to associate at least one logical proposition of the plurality of logical propositions with the at least one non-normalized syntactic structure;

receiving the association between the at least one logical proposition and the at least one non-normalized syntactic structure;

storing the association in the at least one meaning loaded entity; and

applying the at least one expert rule to at least one document to associate the at least one document with the at least one meaning loaded entity.

14. The method according to claim 13, wherein providing, by a computer, an input component that allows the user to provide at least one non-normalized syntactic structure comprises providing, to the user in a display, an input component that allows the user to enter at least one of a word, a phrase and a sentence.

15. The method according to claim 13, further comprising

providing an input component that allows the user to provide at least one meaning taxonomy element;

receiving the at least one meaning taxonomy element;

providing an input component that allows the user to associate the at least one meaning taxonomy element with the at least one meaning loaded entity; and

receiving the association between the at least one meaning taxonomy element and the at least one meaning loaded entity.

16. A system for building associations between a meaning taxonomy and at least one document comprising:

a first input configured to receive the at least one document from another system, the at least one document including syntactic structures;

a second input configured to receive a meaning taxonomy including a plurality of selected concepts and a group of syntactic structures including at least one syntactic structure, at least one of the group of syntactic structures being associated with at least one selected concept of the plurality of selected concepts;

a third input configured to receive a set of expert rules associated with the at least one selected concept, each expert rule of the set of expert rules including a plurality of logical propositions, at least one logical proposition of the plurality of logical propositions including an evaluation of whether a relationship an association exists between one or more of the group of syntactic structures associated with the at least one selected concept and one or more syntactic structures included in the at least one document;

a memory having storage capacity; and

a controller coupled to the memory and the first, second and third inputs and configured to store, responsive to existence of the association, an association of the at least one document with the meaning taxonomy in the memory.

17. The system according to claim 16, wherein the plurality of selected concepts includes a plurality of meaning loaded entities.

18. The system according to claim 16, wherein the set of expert rules includes a set of logical propositions, the logical propositions being expressed in terms of meaning loaded entities.

19. The system according to claim 16, wherein the controller is further configured to store an association of the at least one document with a selected concept within the meaning taxonomy, the selected concept being associated with a non-normalized syntactic structure that is included in the at least one document.

20. The system according to claim 16, wherein the controller is further configured to store an association of the at least one document with a meaning loaded entity within the meaning taxonomy, the meaning loaded entity being associated with an expert rule including a logical proposition that evaluates to true when evaluated against the at least one document.