Method and system for acquisition, normalization, matching, and enrichment of data

• US 10,223,410 B2
• Filed: 08/31/2014
• Issued: 03/05/2019
• Est. Priority Date: 01/06/2014
• Status: Active Grant

• Current Assignee: Cisco Technology, Inc. (Cisco Systems, Inc.)
• Original Assignee: Cisco Technology, Inc. (Cisco Systems, Inc.)
• Inventors: Tallamy, Stephen Charles, Bertrand, Laurent Edmond Pierre, Rai, Pete
• Primary Examiner(s): Burke, Jeff A
• Assistant Examiner(s): Vu, Thong H
• Application Number:
  US14/474,153
• Publication Number:
  US 20150193497A1
• Time in Patent Office: 1,647 Days
• Field of Search: 707770, 707724, 707728, 707748, 707759, 707722, 707600, 707769, 705 30
• CPC Class Codes: G06F 16/24 Querying
• CPC Class Codes: G06N 5/02 Knowledge representation; Symbolic representation
• CPC Class Codes: G06N 5/022 Knowledge engineering; Knowledge acquisition

Abstract

In one embodiment, a method includes obtaining a first data set from a first data source and a second data set from a second data source, the first data set including a first plurality of entities and the second data set including a second plurality of entities. The method also includes identifying a verified relationship between a first entity from the first plurality of entities and a second entity from the second plurality of entities and determining that a third entity from the first plurality of entities has a first same-as relationship with a fourth entity from the second plurality of entities based on one or more of the verified relationship or relationships between the first plurality of entities and the second plurality of entities. The method further includes generating first output data including the first same-as relationship.

Claims

1. A method comprising:

obtaining a first data set from a first data source and a second data set from a second data source, the first data set comprising a first plurality of entities and the second data set comprising a second plurality of entities;

identifying a verified same-as relationship between a first entity from the first plurality of entities and a second entity from the second plurality of entities;

comparing one or more entities of the first plurality of entities having respective indicated relationships with the first entity with one or more entities of the second plurality of entities having respective indicated relationships with the second entity;

determining that a third entity from the first plurality of entities has an inferred same-as relationship with a fourth entity from the second plurality of entities based on comparing the one or more entities of the first plurality of entities having respective indicated relationships with the first entity with the one or more entities of the second plurality of entities having respective indicated relationships with the second entity;

comparing one or more entities of the first plurality of entities having respective indicated relationships with the third entity with one or more entities of the second plurality of entities having respective indicated relationships with the fourth entity;

classifying the inferred same-as relationship into one of a plurality of accuracy classes based on comparing the one or more entities of the first plurality of entities having respective indicated relationships with the third entity with the one or more entities of the second plurality of entities having respective indicated relationships with the fourth entity; and

generating first output data indicating that the third entity has the classified inferred same-as relationship with the fourth entity.

2. The method of claim 1, wherein identifying the verified same-as relationship between the first entity and the second entity comprises:

receiving data indicative of the verified same-as relationship.

3. The method of claim 1, wherein classifying the inferred same-as relationship comprises classifying the inferred same-as relationship as either a candidate same-as relationship or a qualified same-as relationship.

4. The method of claim 1, wherein classifying the inferred same-as relationship is based on a number of the one or more entities of the first plurality of entities having respective indicated relationships with the third entity that have respective same-as relationships with the one or more entities of the second plurality of entities having respective indicated relationships with the fourth entity.

5. The method of claim 4, wherein the inferred same-as relationship is classified as a qualified same-as relationship when the number exceeds a threshold number.

6. The method of claim 1, further comprising:

identifying one or more of an inferred relationship or an inferred attribute of a fifth entity based on one or more of a relationship between entities of the first plurality of entities and the second plurality of entities or attributes of the first plurality of entities and the second plurality of entities.

7. The method of claim 1, further comprising:

providing the first output data to an information system for naming, describing, or associating entities in the information system.

8. The method of claim 6, further comprising:

generating second output data indicating one or more of the inferred relationship or the inferred attribute.

9. The method of claim 8, further comprising:

generating third output data based on the first output data and the second output data, the third output data comprising rights-free data.

10. The method of claim 1, further comprising:

receiving a plurality of data sets from a plurality of data providers, a plurality of matcher modules, and a plurality of inference modules, the plurality of data sets comprising same-as relationships identified by the plurality of matcher modules and inferred data identified by the plurality of inference modules;

receiving an indication of a trust level, the trust level indicative of one or more of allowed data providers, allowed matcher modules, allowed inference modules, allowed relationship qualifiers, or allowed attribute qualifiers;

identifying a first subset of the plurality of data sets that satisfy the trust level; and

generating an output data set based on the first subset, the output data set comprising a second subset of same-as relationships and a third subset of the inferred data.

11. The method of claim 1, further comprising:

receiving a plurality of data sets from a plurality of data providers, a plurality of matcher modules, and a plurality of inference modules, the plurality of data sets comprising conflicting data about one or more of an entity, an attribute, or a relationship between entities, further comprising same-as relationships identified by the plurality of matcher modules, and further comprising inferred data identified by the plurality of inference modules;

determining whether a first data set from the plurality of data sets identifies entities, attributes, and relationships using a common set of predicates; and

modifying the first data set to identify entities, attributes, and relationships using the common set of predicates when the first data set does not use the common set of predicates.

12. A system comprising:

a processor; and

a combination of logic and non-transitory memory including instructions, that when at least executed in part by the processor cause the system to;

obtain a first data set from a first data source and a second data set from a second data source, the first data set comprising a first plurality of entities and the second data set comprising a second plurality of entities;

identify a verified same-as relationship between a first entity from the first plurality of entities and a second entity from the second plurality of entities;

compare one or more entities of the first plurality of entities having respective indicated relationships with the first entity with one or more entities of the second plurality of entities having respective indicated relationships with the second entity;

determine that a third entity from the first plurality of entities has an inferred same-as relationship with a fourth entity from the second plurality of entities based on comparing the one or more entities of the first plurality of entities having respective indicated relationships with the first entity with the one or more entities of the second plurality of entities having respective indicated relationships with the second entity;

compare one or more entities of the first plurality of entities having respective indicated relationships with the third entity with one or more entities of the second plurality of entities having respective indicated relationships with the fourth entity;

classify the inferred same-as relationship into one of a plurality of accuracy classes based on comparing the one or more entities of the first plurality of entities having respective indicated relationships with the third entity with the one or more entities of the second plurality of entities having respective indicated relationships with the fourth entity; and

generate first output data indicating that the third entity has the classified inferred same-as relationship with the fourth entity.

13. The system of claim 12, wherein classifying the inferred same-as relationship comprises classifying the inferred same-as relationship as either a candidate same-as relationship or a qualified same-as relationship.

14. The system of claim 12, wherein classifying the inferred same-as relationship is based on a number of the one or more entities of the first plurality of entities having respective indicated relationships with the third entity that have respective same-as relationships with the one or more entities of the second plurality of entities having respective indicated relationships with the fourth entity.

15. The system of claim 14, wherein the inferred same-as relationship is classified as a qualified same-as relationship when the number exceeds a threshold number.

16. The system of claim 12, wherein the combination of logic and non-transitory memory further cause the processor to:

provide the first output data to an information system for naming, describing, or associating entities for TV or video, social media, online shopping, or web content.

17. The system of claim 12, wherein the combination of logic and non-transitory memory further cause the processor to:

identify one or more of an inferred relationship or an inferred attribute of a fifth entity based on one or more of a relationship between entities of the first plurality of entities and the second plurality of entities or attributes of the first plurality of entities and the second plurality of entities.

18. The system of claim 12, wherein the combination of logic and non-transitory memory further cause the processor to:

generate second output data indicating one or more of the inferred relationship or the inferred attribute.

19. The system of claim 18, wherein the combination of logic and non-transitory memory further cause the processor to:

generate third output data based on the first output data and the second output data, the third output data comprising rights-free data.