Method For Deducing Entity Relationships Across Corpora Using Cluster Based Dictionary Vocabulary Lexicon

• US 20170255694A1
• Filed: 05/17/2017
• Published: 09/07/2017
• Est. Priority Date: 09/26/2014
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Byron, Donna K., Chandrasekaran, Swaminathan, Krishnamurthy, Lakshminarayanan
• Granted Patent: US 10,664,505 B2
• Time in Patent Office: Days
• CPC Class Codes: G06F 16/245 Query processing
• CPC Class Codes: G06F 16/3344 using natural language analysis
• CPC Class Codes: G06F 16/355 Class or cluster creation or modification
• CPC Class Codes: G06F 40/30 Semantic analysis

Abstract

An approach is provided for identifying entity relationships based on word classifications extracted from business documents stored in a plurality of corpora. In the approach, performed by an information handling system, a plurality of cluster classifications are identified for the business documents so that entity information from the business documents can be classified or assigned to the cluster classifications, such as by performing natural language processing (NLP) analysis of the business documents. The approach applies semantic analysis to identify and score entity relationships between the entity information classified in the cluster classifications, and based on the scored entity relationships, cluster relationships between the cluster classifications are identified.

Claims

1. A method, in an information handling system comprising a processor and a memory, of identifying cluster relationships for searching across a plurality of corpora, the method comprising:

identifying, by the system, a plurality of different cluster classifications for a corresponding plurality of corpora;

classifying, by the system, entity information from documents stored in the plurality of corpora into the plurality of different cluster classifications;

applying semantic analysis, by the system, to identify entity relationships between entity information classified in the plurality of different cluster classifications;

determining, by the system, one or more scores for each identified entity relationship;

identifying, by the system, a cluster relationship between at least two cluster classifications based on the one or more scores for each identified entity relationship; and

searching, by the information handling system, at least first and second corpora corresponding to the at least two cluster classifications having the identified cluster relationship.

2. The method of claim 1, wherein classifying entity information from the documents comprises performing, by the system, a natural language processing (NLP) analysis of the documents, wherein the NLP analysis extracts named entity information from the documents.

3. The method of claim 1, wherein the documents are selected from a group comprising billing, customer orders, procedures, dealers, customer correspondence, credit, incidence, and service documents.

6. The method of claim 1, wherein determining one or more scores comprises applying a normalized weighting algorithm to the one or more scores to generate normalized scores for each identified entity relationship.

7. The method of claim 6, further comprising applying a threshold test to the normalized scores to identify each entity relationship having a normalized score that meets or exceeds the threshold test.

8. The method of claim 7, further comprising constructing a model which specifies cluster relationships within the plurality of different cluster classifications.

4-5. -5. (canceled)

9. An information handling system comprising:

one or more processors;

a memory coupled to at least one of the processors;

a set of instructions stored in the memory and executed by at least one of the processors to identify cluster relationships for searching across a plurality of corpora, wherein the set of instructions perform actions of;

identifying, by the system, a plurality of different cluster classifications for corresponding plurality of corpora;

classifying, by the system, entity information from documents stored in the plurality of corpora into the plurality of different cluster classifications;

applying semantic analysis, by the system, to identify entity relationships between entity information classified in the plurality of different cluster classifications;

determining, by the system, one or more scores for each identified entity relationship;

identifying, by the system, a cluster relationship between at least two cluster classifications based on the one or more scores for each identified entity relationships; and

searching, by the information handling system, at least first and second corpora corresponding to the at least two cluster classifications having the identified cluster relationship.

10. The information handling system of claim 9, wherein classifying entity information from the documents comprises performing, by the system, a natural language processing (NLP) analysis of the documents, wherein the NLP analysis extracts named entity information from the documents.

11. The information handling system of claim 9, wherein the documents are selected from a group comprising billing, customer orders, procedures, dealers, customer correspondence, credit, incidence, and service documents.

14. The information handling system of claim 9, wherein determining one or more scores comprises applying a normalized weighting algorithm to the one or more scores to generate normalized scores for each identified entity relationship.

15. The information handling system of claim 14, further comprising applying a threshold test to the normalized scores to identify each identified entity relationship having a normalized score that meets or exceeds the threshold test.

16. The information handling system of claim 15, wherein the actions further comprise constructing a model which specifies cluster relationships within the plurality of different cluster classifications.

12-13. -13. (canceled)

17. A computer program product stored in a computer readable storage medium, comprising computer instructions that, when executed by an information handling system, causes the information handling system to identify cluster relationships for searching across a plurality of corpora by performing actions comprising:

identifying, by the system, a plurality of different cluster classifications for a corresponding plurality of corpora;

classifying, by the system, entity information from documents stored in the plurality of corpora into the plurality of different cluster classifications;

applying semantic analysis, by the system, to identify entity relationships between entity information classified in the plurality of different cluster classifications;

determining, by the system, one or more scores for each identified entity relationship;

identifying, by the system, a cluster relationship between at least two cluster classifications based on the one or more scores for each identified entity relationships; and

searching, by the information handling system, at least first and second corpora corresponding to the at least two cluster classifications having the identified cluster relationship.

20. The computer program product of claim 17, wherein determining one or more scores comprises applying a normalized weighting algorithm to the one or more scores to generate normalized scores for each identified entity relationship, and applying a threshold test to the normalized scores to identify each entity relationship having a normalized score that meets or exceeds the threshold test.

18-19. -19. (canceled)