Method and system for identifying entities

• US 9,965,508 B1
• Filed: 06/20/2014
• Issued: 05/08/2018
• Est. Priority Date: 10/14/2011
• Status: Active Grant

• Current Assignee: Aurea Software Incorporated (ESW Capital, LLC)
• Original Assignee: Ignite Firstrain Solutions, Inc.
• Inventors: Cooke, David, Betz, Martin, Joshi, Ashutosh, Mohanty, Binay
• Primary Examiner(s): Moser, Bruce
• Application Number:
  US14/311,179
• Time in Patent Office: 1,418 Days
• CPC Class Codes: G06F 16/245 Query processing
• CPC Class Codes: G06F 16/26 Visual data mining; Browsing structured data
• CPC Class Codes: G06F 16/35 Clustering; Classification
• CPC Class Codes: G06F 16/353 into predefined classes
• CPC Class Codes: G06F 16/93 Document management systems
• CPC Class Codes: G06F 16/9535 Search customisation based on user profiles and personalisation
• CPC Class Codes: G06F 40/30 Semantic analysis

Abstract

Some embodiments provide a program that identifies an entity having an entity attribute. The program receives, from each method of several methods, a set of candidate identity attributes that are each for identifying a particular entity having the entity attribute specified in the document. Each method of the several methods generates the corresponding set of candidate identity attributes based on the entity attribute specified in a document. The program calculates a score for each candidate identity attribute in the sets of candidate identity attributes. The program identifies, based on the sets of scores, an identity attribute from the sets of candidate identity attributes that identifies the entity having the entity attribute specified in the document.

Claims

1. A non-transitory machine readable medium storing a program which when executed by at least one processing unit identifies a set of identity attributes for determining the identity of an entity, the program comprising sets of instructions for:

identifying a particular name that occurs more often than other names in a set of documents;

identifying a plurality of candidate identity attribute sets by analyzing the particular name and at least one document in the set of documents using a plurality of different processes that each identifies (i) a set of candidate identities corresponding to the particular name and (ii) a candidate identity attribute set for each identified candidate identity, wherein at least one of the different processes analyzes a stored plurality of identities to identify candidate identities having the particular name and that are related to an entity to which the at least one document is also related;

for each candidate identity attribute set of the plurality of candidate identity attribute sets, calculating a relevance score for each candidate identity attribute in the set that measures a level of correspondence between the particular name and the candidate identity attribute; and

identifying, based on the relevance scores calculated for the candidate identity attributes of the different candidate identity attribute sets, a particular candidate identity attribute set for a particular identity that corresponds to the particular name.

2. The non-transitory machine readable medium of claim 1, wherein the set of instructions for identifying the plurality of candidate identity attribute sets comprises a set of instructions for querying a set of databases to retrieve a candidate identity and corresponding candidate identity attribute set based on the particular name.

3. The non-transitory machine readable medium of claim 1, wherein each candidate identity attribute set comprises a first candidate identity attribute of a first type and a second candidate identity attribute of a second type.

4. The non-transitory machine readable medium of claim 3, wherein the particular name is a name of a person, the first type is a title of the person, and the second type is a company with which the person is affiliated.

5. The non-transitory machine readable medium of claim 3, wherein the set of instructions for calculating the relevance scores comprises sets of instructions for:

processing the plurality of candidate identity attribute sets based on the first candidate identity attribute of each candidate identity attribute set to identify a subset of the plurality of candidate identity attribute sets; and

processing only the subset of the plurality of candidate identity attribute sets based on the second candidate identity attribute of each candidate identity attribute set.

6. The non-transitory machine readable medium of claim 1, wherein the set of instructions for identifying the particular candidate identity attribute set comprises a set of instructions for identifying the candidate identity attribute set of the plurality of candidate identity attribute sets with a highest set of relevance scores.

7. The non-transitory machine readable medium of claim 1, wherein the at least one document is categorized as related to a particular industry, wherein a particular one of the processes analyzes the stored plurality of identities to identify a candidate identity that has the particular name and that is associated with a company in the particular industry.

8. The non-transitory machine readable medium of claim 1, wherein a particular one of the processes analyzes the stored plurality of identities to identify a candidate identity that has the particular name and that is associated with a company mentioned in the at least one document.

9. A method for identifying a set of identity attributes for determining the identity of an entity, the method comprising:

identifying a particular name that occurs more often than other names in a set of documents;

identifying a plurality of candidate identity attribute sets by analyzing the particular name and at least one document in the set of documents using a plurality of different processes that each identifies (i) a set of candidate identities corresponding to the particular name and (ii) a candidate identity attribute set for each identified candidate identity, wherein at least one of the different processes analyzes a stored plurality of identities to identify candidate identities having the particular name and that are related to an entity to which the at least one document is also related;

for each candidate identity attribute set of the plurality of candidate identity attribute sets, calculating a relevance score for each candidate identity attribute in the set that measures a level of correspondence between the particular name and the candidate identity attribute; and

identifying, based on the relevance scores calculated for the candidate identity attributes of the different candidate identity attribute sets, a particular candidate identity attribute set for a particular identity that corresponds to the particular name.

10. The method of claim 9, wherein identifying the plurality of candidate identity attribute sets comprises identifying a candidate identity attribute set based on a lexical analysis of the at least one document.

11. The method of claim 9 further comprising calculating a normalization factor that converts a particular relevance score calculated for a particular candidate identity attribute to a particular range of values.

12. The method of claim 9, wherein calculating a particular relevance score for a particular candidate identity attribute for a particular candidate identity comprises calculating a confidence factor that represents a probability that the particular candidate identity correctly identifies the entity referred to by the particular name in the set of documents.

13. The method of claim 9, wherein each candidate identity attribute set comprises a first candidate identity attribute of a first type and a second candidate identity attribute of a second type.

14. The method of claim 13, wherein the particular name is a name of a person, the first type is a title of the person, and the second type is a company with which the person is affiliated.

15. The method of claim 13, wherein calculating the relevance scores comprises:

processing the plurality of candidate identity attribute sets based on the first candidate identity attribute of each candidate identity attribute set to identify a subset of the plurality of candidate identity attribute sets; and

processing only the subset of the plurality of candidate identity attribute sets based on the second candidate identity attribute of each candidate identity attribute set.

16. A device comprising:

a set of processing units; and

a non-transitory machine readable medium storing a program which when executed by the set of processing units determines the identity of an entity, the program comprising sets of instructions for;

identifying a particular name that occurs more often than other names in a set of documents;

identifying a plurality of candidate identity attribute sets by analyzing the particular name and at least one document in the set of documents using a plurality of different processes that each identifies (i) a set of candidate identities corresponding to the particular name and (ii) a candidate identity attribute set for each identified candidate identity, wherein at least one of the different processes analyzes a stored plurality of identities to identify candidate identities having the particular name and that are related to an entity to which the at least one document is also related;

for each candidate identity attribute set of the plurality of candidate identity attribute sets, calculating a relevance score for each candidate identity attribute in the set that measures a level of correspondence between the particular name and the candidate identity attribute; and

identifying, based on the relevance scores calculated for the candidate identity attributes of the different candidate identity attribute sets, a particular candidate identity attribute set for a particular identity that corresponds to the particular name.

17. The device of claim 16, wherein the set of instructions for identifying the plurality of candidate identity attribute sets comprises a set of instructions for querying a set of databases to identify candidate identities and corresponding candidate identity attribute sets based on the particular name.

18. The device of claim 16, wherein the set of instructions for identifying the plurality of candidate identity attribute sets comprises a set of instructions for performing a lexical analysis of the at least one document to identify candidate identities and corresponding candidate identity attribute sets.

19. The device of claim 16, wherein the program further comprises a set of instructions for calculating a normalization factor that converts a particular relevance score calculated for a particular candidate identity attribute to a particular range of values.

20. The device of claim 16, wherein the set of instructions for identifying the particular candidate identity attribute set comprises a set of instructions for identifying the candidate identity attribute set of the plurality of candidate identity attribute sets with a highest set of relevance scores.