Methods and apparatus for entity detection

• US 9,129,013 B2
• Filed: 03/12/2013
• Issued: 09/08/2015
• Est. Priority Date: 03/12/2013
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Nuance Communications, Inc. (Microsoft Corporation)
• Inventors: Delaney, Brian W., Yegnanarayanan, Girija
• Primary Examiner(s): MOSER, BRUCE M
• Application Number:
  US13/795,886
• Publication Number:
  US 20140280353A1
• Time in Patent Office: 910 Days
• US Class Current: 1/1
• CPC Class Codes: G06F 16/367 Ontology
• CPC Class Codes: G06F 40/279 Recognition of textual entities
• CPC Class Codes: G06F 40/295 Named entity recognition
• CPC Class Codes: G06F 40/30 Semantic analysis
• CPC Class Codes: G16H 10/60 for patient-specific data, e.g. for electronic patient records

Abstract

Techniques for entity detection include matching a token from at least a portion of a text string with a matching concept in an ontology. A first concept may be identified as being hierarchically related to the matching concept within the ontology, and a second concept may be identified as being hierarchically related to the first concept within the ontology. The first and second concepts may be included in a set of features of the token. Based at least in part on the set of features of the token, a measure related to a likelihood that the at least a portion of the text string corresponds to a particular entity type may be determined.

Claims

1. A method comprising:

matching a token from at least a portion of a text string with a matching concept in an ontology;

identifying a first concept as being hierarchically related to the matching concept within the ontology;

identifying a second concept as being hierarchically related to the first concept within the ontology;

including the first and second concepts in a set of features of the token; and

determining, using at least one processor, a measure related to a likelihood that the at least a portion of the text string corresponds to a particular entity type, based at least in part on the set of features of the token.

2. The method of claim 1, wherein the particular entity type is a particular type of medical fact.

3. The method of claim 1, wherein the first concept is an ancestor of the matching concept within the ontology.

4. The method of claim 3, wherein the second concept is an ancestor of the first concept within the ontology.

5. The method of claim 4, wherein the first concept is a parent concept of the matching concept within the ontology, and wherein the second concept is a parent concept of the first concept within the ontology.

6. The method of claim 1, wherein the matching concept represents a semantic meaning of the token.

7. The method of claim 1, wherein the determining comprises inputting the first and second concepts, as features of the token, to a statistical model trained to determine the measure based on a plurality of features of the token.

8. The method of claim 1, wherein the determining comprises weighting the first and second concepts as features of the token relative to other features of the token.

9. Apparatus comprising:

at least one processor; and

at least one processor-readable storage medium storing processor-executable instructions that, when executed by the at least one processor, perform a method comprising;

matching a token from at least a portion of a text string with a matching concept in an ontology;

identifying a first concept as being hierarchically related to the matching concept within the ontology;

identifying a second concept as being hierarchically related to the first concept within the ontology;

including the first and second concepts in a set of features of the token; and

determining a measure related to a likelihood that the at least a portion of the text string corresponds to a particular entity type, based at least in part on the set of features of the token.

10. The apparatus of claim 9, wherein the particular entity type is a particular type of medical fact.

11. The apparatus of claim 9, wherein the first concept is a parent concept of the matching concept within the ontology, and wherein the second concept is a parent concept of the first concept within the ontology.

12. The apparatus of claim 9, wherein the determining comprises inputting the first and second concepts, as features of the token, to a statistical model trained to determine the measure based on a plurality of features of the token.

13. At least one computer-readable storage medium encoded with computer-executable instructions that, when executed, perform a method comprising:

matching a token from at least a portion of a text string with a matching concept in an ontology;

identifying a first concept as being hierarchically related to the matching concept within the ontology;

identifying a second concept as being hierarchically related to the first concept within the ontology;

including the first and second concepts in a set of features of the token; and

determining a measure related to a likelihood that the at least a portion of the text string corresponds to a particular entity type, based at least in part on the set of features of the token.

14. The at least one computer-readable storage medium of claim 13, wherein the particular entity type is a particular type of medical fact.

15. The at least one computer-readable storage medium of claim 13, wherein the first concept is an ancestor of the matching concept within the ontology.

16. The at least one computer-readable storage medium of claim 15, wherein the second concept is an ancestor of the first concept within the ontology.

17. The at least one computer-readable storage medium of claim 16, wherein the first concept is a parent concept of the matching concept within the ontology, and wherein the second concept is a parent concept of the first concept within the ontology.

18. The at least one computer-readable storage medium of claim 13, wherein the matching concept represents a semantic meaning of the token.

19. The at least one computer-readable storage medium of claim 13, wherein the determining comprises inputting the first and second concepts, as features of the token, to a statistical model trained to determine the measure based on a plurality of features of the token.

20. The at least one computer-readable storage medium of claim 13, wherein the determining comprises weighting the first and second concepts as features of the token relative to other features of the token.