Semi-supervised identity aggregation of profiles using statistical methods

Patent metadata

Field	Value
Publication number	US-9654594-B2
Application number	US-201313932269-A
Country	US
Kind code	B2
Filing date	Jul 1, 2013
Priority date	Sep 10, 2012
Publication date	May 16, 2017
Grant date	May 16, 2017

Abstract

User profiles can be analyzed to identify profiles matching to the same identity. For example, profiles from different social network systems are analyzed to determine if the profiles are associated with the same user of the social network systems. Multiple heuristics may be calculated using different algorithms. The calculated heuristics may then be combined to generate a match score that indicates whether two profiles match.

First claim

What is claimed is:

1. A method for aggregating social network profiles, the method comprising:
accessing, by a talent management system of an enterprise, data of a first profile from a first data source, wherein the first data source comprises at least one internal data source operated by the enterprise and wherein the first profile comprises a profile of an employee of the enterprise or candidate for employment by the enterprise and includes human resource information of the enterprise related to the employee of the enterprise or candidate for employment by the enterprise;
accessing, by the talent management system, data of a second profile from a second data source, wherein the second data source comprises at least one external data source operated by or on behalf of an entity different from the enterprise, the at least one external data source comprising at least one social network system;
determining, by the talent management system, whether the first profile contains a uniform resource locator (URL) that identifies the second profile;
in response to determining that the first profile contains the URL that identifies the second profile, identifying, by the talent management system, the second profile as matching the first profile;
in response to determining that the first profile does not contain the URL that identifies the second profile:
calculating, by the talent management system, one or more heuristics, each of the one or more heuristics being based on one or more attributes of the first profile having a matching one or more attributes in the second profile,
weighting, by the talent management system, the one or more heuristics based on, for each of the one or more heuristics, a reliability of the one or more attributes on which the heuristic is based,
calculating, by the talent management system, a match score from the one or more weighted heuristics; and
determining, by the talent management system, whether the first profile and the second profile match based on the calculated match score; and

in response to determining that the first profile and the second profile match, storing an indication that the first profile and the second profile match and the second profile belongs to the employee of the enterprise or candidate for employment by the enterprise.

2. The method of claim 1 , wherein calculating one of the one or more heuristics comprises:
obtaining attributes of the first profile;
generating attributes of the second profile;
identifying attributes of the second profile matching attributes of the first profile; and
calculating the one of the one or more heuristics based at least in part on the identified matching attributes, and wherein weighting the one of the one or more heuristics is further based a weight that indicates a significance of the one of the one or more heuristics and is calculated based on an accuracy of matches made in examined social profiles.

3. The method of claim 1 , wherein calculating one of the one or more heuristics comprises:
obtaining one or more expressions of interest from the first profile;
obtaining one or more expressions of interest from the second profile;
identifying expressions of interest from the second profile matching expressions of interest from the first profile; and
calculating the one or more heuristics based at least in part on the identified matching expressions of interest, and wherein weighting the one of the one or more heuristics is further based on a weight that indicates a significance of the one of the one or more heuristics and is calculated based on an accuracy of matches made in examined social profiles.

4. The method of claim 1 , wherein calculating one of the one or more heuristics comprises:
obtaining one or more profile connections from the first profile;
obtaining one or more profile connections from the second profile;
identifying profile connections from the second profile matching profile connections from the first profile; and
calculating the one or more heuristics based at least in part on the identified matching profile connections, and wherein weighting the one of the one or more heuristics is further based on a weight that indicates a significance of the one of the one or more heuristics and is calculated based on an accuracy of matches made in examined social profiles.

5. The method of claim 1 , wherein calculating one of the one or more heuristics comprises:
obtaining one or more images associated with the first profile;
obtaining one or more images associated with the second profile;
identifying faces within the obtained images associated with the first profile and the images associated with the second profile;
identifying images associated with the second profile matching images associated with the first profile based on the identified faces; and
calculating the one or more heuristics based at least in part on the identified matching images, and wherein weighting the one of the one or more heuristics is further based on a weight that indicates a significance of the one of the one or more heuristics and is calculated based on an accuracy of matches made in examined social profiles.

6. The method of claim 1 , wherein calculating one of the one or more heuristics comprises:
obtaining content posted in connection with the first profile;
obtaining content posted in connection with the second profile;
identifying content posted in connection with the first profile that at least partially matches content posted in connection with the second profile;
calculating a number of instances of at least partially matching content that is posted to both the first profile and the second profile within a time threshold; and
calculating the one or more heuristics based at least in part on the calculated number of instances, and wherein weighting the one of the one or more heuristics is further based on a weight that indicates a significance of the one of the one or more heuristics and is calculated based on an accuracy of matches made in examined social profiles.

7. The method of claim 1 , wherein calculating one of the one or more heuristics comprises:
obtaining a set of hyperlinks posted in connection with the first profile;
obtaining a set of hyperlinks posted in connection with the second profile;
identifying hyperlinks posted in connection with the first profile that at least partially matches hyperlinks posted in connection with the second profile;
calculating a number of matching hyperlinks that are posted to both the first profile and the second profile; and
calculating the one or more heuristics based at least in part on the calculated number of matching hyperlinks, and wherein weighting the one of the one or more heuristics is further based on a weight that indicates a significance of the one of the one or more heuristics and is calculated based on an accuracy of matches made in examined social profiles.

8. A system comprising:
a processor; and
a memory coupled with and readable by the processor and storing therein a set of instructions which, when executed by the processor, causes the processor to aggregate social network profiles by:
accessing data of a first profile from a first data source, wherein the first data source comprises at least one internal data source operated by an enterprise and wherein the first profile comprises a profile of an employee of the enterprise or candidate for employment by the enterprise and includes human resource information of the enterprise related to the employee of the enterprise or candidate for employment by the enterprise;
accessing data of a second profile from a second data source, wherein the second data source comprises at least one external data source operated by or on behalf of a

Assignees

• Oracle Int Corp

Inventors

• B'Far Reza
• Spaulding Kent
• Cengiz Yasin
• Caves Americo
• Goh Fun
• Lingg Elizabeth
• Kal Yenal
• Shi Yemin
• Fu Chengxiao

Classifications

• G06Q10/40

  Business processes related to social networking or social networking services · CPC title

• H04L67/306 Primary

  User profiles · CPC title

• H04L63/20

  for managing network security; network security policies in general (filtering policies H04L63/0227) · CPC title

• H04L63/102

  Entity profiles · CPC title

• G06Q50/01

  Physics · mapped topic