Discovery of related entities in a master data management system

Patent metadata

Field	Value
Publication number	US-10042911-B2
Application number	US-201313954155-A
Country	US
Kind code	B2
Filing date	Jul 30, 2013
Priority date	Jul 30, 2013
Publication date	Aug 7, 2018
Grant date	Aug 7, 2018

Abstract

Methods and arrangements for discovering entity types for a set of records. A set of records is input, with each record comprising attributes with associated attribute values. The records are grouped into candidate entity types in view of at least one of: the attribute values of the records, at least one domain ontology and at least one dimension hierarchy. An interestingness measure of each candidate entity type is calculated, via estimating interestingness based on at least one factor selected from the group consisting of: a correlation between attribute values of records, a number of attributes, a log of queries issued to a server, and an average group size for candidate entity types. At least one candidate entity type is validated based on the calculated interestingness measures. Other variants and embodiments are broadly contemplated herein.

First claim

What is claimed is:

1. A method of discovering entity types for a set of records, said method comprising:
inputting a set of records, each record comprising attributes with associated attribute values;
grouping the records into candidate entity types in view of at least one of: the attribute values of the records, at least one domain ontology, and at least one dimension hierarchy, wherein the grouping comprises constructing a lattice space of attribute combinations and wherein the candidate entity types are based upon one or more of the attribute combinations;
calculating an interestingness measure of each candidate entity type being associated with one or more of the attribute combinations, wherein the interestingness measure comprises a measure of relevance of a candidate entity type and wherein the calculating comprises estimating interestingness of the one or more attribute combinations associated with the candidate entity type based on a correlation between attribute values of records and attributing the interestingness of the one or more attribute combination associated with the candidate entity type to the candidate entity type;
the correlation being identified by calculating a correlation score between a pair of attributes of an attribute combination of the candidate entity type, wherein the correlation score is based upon functional dependency between the pair of attributes and an intended usage of the candidate entity type;
ranking the candidate entity types based upon the interestingness measure attributed to each candidate entity types;
validating, by ranked order, at least one of the candidate entity types based on the calculated interestingness measures; and
grouping, using the validated candidate entity types, attributes of entities within the set of records into entity types identified by the validated candidate entity types.

2. The method according to claim 1 , wherein:
said validating comprises assisting a user in validating at least one candidate entity type; and
assisting a user in creating at least one new candidate entity type.

3. The method according to claim 2 , wherein said validating comprises determining a relevance of each candidate entity type to at least one preconfigured entity type and presenting the candidate entity types to the user in an order of relevance.

4. The method according to claim 2 , wherein said calculating comprises:
employing a plurality of individual measures to determine interestingness;
assigning relative weights to the individual measures; and
combining the weighted individual measures.

5. The method according to claim 1 , comprising establishing relationships between a plurality of candidate entity types.

6. The method according to claim 5 , wherein said establishing of relationships comprises linking discovered entity types via discovering an is-a relationship.

7. The method according to claim 6 , wherein said establishing of relationships comprises linking discovered entity types additionally via discovering one or more sibling relationships.

8. The method according to claim 1 , wherein said grouping comprises consulting all of: the attribute values of the records, domain ontologies and dimension hierarchies.

9. The method according to claim 8 , wherein said grouping comprises identifying candidate entity types at one or more hierarchy levels via using attributes relevant to each candidate entity type.

10. An apparatus for discovering entity types for a set of records, said apparatus comprising:
at least one hardware processor; and
a computer readable storage medium having computer readable program code embodied therewith and executable by the at least one processor, the computer readable program code comprising:
computer readable program code configured to input a set of records, each record comprising attributes with associated attribute values;
computer readable program code configured to group the records into candidate entity types in view of at least one of: the attribute values of the records, at least one domain ontology and at least one dimension hierarchy, wherein the grouping comprises constructing a lattice space of attribute combinations and wherein the candidate entity types are based upon one or more of the attribute combinations;
computer readable program code configured to calculate an interestingness measure of each candidate entity type being associated with one or more of the attribute combinations, wherein the interestingness measure comprises a measure of relevance of a candidate entity type and wherein the calculating comprises estimating interestingness of the one or more attribute combinations associated with the candidate entity type based on a correlation between attribute values of records and attributing the interestingness of the one or more attribute combination associated with the candidate entity type to the candidate entity type;
the correlation being identified by calculating a correlation score between a pair of attributes of an attribute combination of the candidate entity type, wherein the correlation score is based upon functional dependency between the pair of attributes and an intended usage of the candidate entity type;
computer readable program code configured to rank the candidate entity types based upon the interestingness measure attributed to each candidate entity types;
computer readable program code configured to validate, by ranked order, at least one of the candidate entity type based on the calculated interestingness measures; and
computer readable program code configured to group, using the validated candidate entity types, attributes of entities within the set of records into entity types identified by the validated candidate entity types.

11. A computer program product for discovering entity types for a set of records, said computer program product comprising:
a computer readable storage medium having computer readable program code embodied therewith, the computer readable program code comprising:
computer readable program code configured to input a set of records, each record comprising attributes with associated attribute values;
computer readable program code configured to group the records into candidate entity types in view of at least one of: the attribute values of the records, at least one domain ontology and at least one dimension hierarchy, wherein the grouping comprises constructing a lattice space of attribute combinations and wherein the candidate entity types are based upon one or more of the attribute combinations;
computer readable program code configured to calculate an interestingness measure of each candidate entity type being associated with one or more of the attribute combinations, wherein the interestingness measure comprises a measure of relevance of a candidate entity type and wherein the calculating comprises estimating interestingness of the one or more attribute combinations associated with the candidate entity type based on a correlation between attribute values of records and attributing the interestingness of the one or more attribute combination associated with the candidate entity type to the candidate entity type;
the correlation being identified by calculating a correlation score between a pair of attributes of an attribute combination of the candidate entity type, wherein the correlation score is based upon functional dependency between the pair of attributes and an intended usage of the candidate entity type;
computer readable program code configured to rank the candidate entity types based upon the interestingness measure attributed to each candidate entity types;
computer readable program code configured to validate, by ranked

Assignees

• IBM
• Int Business Machines Corporations

Inventors

• Deshpande Prasad M
• Joshi Salil R
• Mohania Mukesh Kumar
• Murthy Karin
• Schumacher Scott
• Sundarmurthy Bruhathi H

Classifications

• G06F16/285 Primary

  Clustering or classification · CPC title

• G06F17/30598 Primary

  Physics · mapped topic