Processes, data structures, and apparatuses for representing knowledge

• US 8,023,739 B2
• Filed: 09/27/2005
• Issued: 09/20/2011
• Est. Priority Date: 09/27/2005
• Status: Expired due to Fees

• Current Assignee: Battelle Memorial Institute
• Original Assignee: Battelle Memorial Institute
• Inventors: Harvey, William J., Thomson, Judi R., Butner, R. Scott, Tratz, Stephen C., Hohimer, Ryan E., Paulson, Patrick R., Chappell, Alan R., Whiting, Mark A.
• Primary Examiner(s): CHANG, JON CARLTON
• Application Number:
  US11/237,483
• Publication Number:
  US 20070083492A1
• Time in Patent Office: 2,184 Days
• Field of Search: 382/181, 382/228, 706/45, 706/46, 707/E17.099
• US Class Current: 382/181
• CPC Class Codes: G06N 5/02 Knowledge representation; Symbolic representation

Abstract

Processes, data structures, and apparatuses to represent knowledge are disclosed. The processes can comprise labeling elements in a knowledge signature according to concepts in an ontology and populating the elements with confidence values. The data structures can comprise knowledge signatures stored on computer-readable media. The knowledge signatures comprise a matrix structure having elements labeled according to concepts in an ontology, wherein the value of the element represents a confidence that the concept is present in an information space. The apparatus can comprise a knowledge representation unit having at least one ontology stored on a computer-readable medium, at least one data-receiving device, and a processor configured to generate knowledge signatures by comparing datasets obtained by the data-receiving devices to the ontologies.

Claims

1. A process for representing knowledge on computational devices, the process comprising:

Providing an ontology stored on a computer-readable medium;

Labeling elements in a knowledge signature according to concepts in the ontology, the knowledge signature comprising a machine-readable data structure derived from the ontology and stored on a computer-readable medium; and

Populating the elements with confidence values by generating an observed confidence value from a data set according to an observation engine and by generating inferred confidence values from at least one observed confidence value according to at least one refinement engine, wherein the confidence values, are calculated by a processor and represent a confidence that the concepts are present in an information space, the observation engine comprises a program module on a computer-readable medium configured to recognize a particular concept within the data set, and the refinement engine comprises a program module on a computer-readable medium configured to infer the presence of concepts related through the ontology to those recognized by the observation engine.

2. The process as recited in claim 1, wherein the knowledge signature is structured as a matrix of the elements.

3. The process as recited in claim 2, Wherein the knowledge signature is a real-valued vector.

4. The process as recited in claim 1, wherein the data set comprises information selected from the group consisting of text documents, video segments, audio segments, images, graphs, database records, sensor data, and combination thereof.

5. The process as recited in claim 1, wherein the observation engine comprises a program module selected from the group consisting of image recognition software, text symbol recognition software, audio recognition software, and combinations thereof.

6. The process as recited in Claim 1, wherein the refinement engine comprises a program module on a computer-readable medium, said program module comprising definitions of concept relationships.

7. The process as recited in claim 6, wherein the concept relationships are selected from the group consisting of subsumption, aggregation, and combinations thereof.

8. The process as recited in claim 1, further comprising generating a concept signature for each concept in the ontology, generating at least one dataset signature from a dataset, and performing a similarity measure between the concept signatures and the dataset signatures to determine confidence values.

9. The process as recited in claim 8, wherein the similarity measure comprises calculating Euclidean distances, cosine coefficients, or combinations thereof.

10. The process as recited in claim 1, further comprising comparing a plurality of knowledge signatures, wherein each knowledge signature is generated from a different data set.

11. The process as recited in claim 10, Wherein the data sets comprise disparate data types.

12. The process as recited in claim 10, wherein said comparing comprises selecting a mode group, transforming the knowledge signatures into reduced-dimensional representations, and performing a similarity measure between the transformed knowledge signatures.

13. The process as recited in claim 12, wherein the similarity measure comprises calculating Euclidian distances, cosine coefficients, or combinations thereof.

14. The process as recited in claim 12, further comprising indexing the reduced-dimensional representation in existing data structures.