Edward Thorne is an Advisory Software Engineer working on InfoSphere Probabilistic Matching Engine for BigInsights in Austin, Texas. Before PME for BigInsights, Ed worked on traditional Master Data Management applications, including InfoSphere Master Data Management and Initiate Master Data Service®. Prior to IBM, Ed worked on enterprise software for retail and content management. Ed graduated from Texas State University with a B.S. in computer science.
InfoSphere Probabilistic Matching Engine for BigInsights (PME for BigInsights) 11.3.0 is the next incremental release of the master data management solution for big data. The PME for BigInsights team released new and enhanced functionality making it easier for developers and administrators to build and maintain their 360-degree information applications.
InfoSphere BigInsights 2.1.2.
Version 11.3.0 was designed for use with InfoSphere BigInsights 2.1.2. This allows PME for BigInsights to leverage features available in Hadoop 2.2.0 and HBase 0.96 that weren't available in BigInsights 2.1.0. Unfortunately it also means customers upgrading from version PME for BigInsights 11.0.0 will also need to upgrade their BigInsights installation.
PME for BigInsights 11.3.0 will now automatically perform entity linking behind the scenes as necessary when data changes. The prior release required scheduling the linking application in the BigInsights console. Just like automatic derivation and comparison, automatic linking can be disabled through table configuration.
You can now request a composited entity view when viewing entity information. This 360-degree view of an entity is assembled using data from the available input sources. By default, the composite view is constructed using the entity's most current attribute values. However, the view can be customized to meet individual requirements.
Entity Data Extraction
Using the new PME Extract application, you can now output entity linkage information as well as composite views of the entities into an HBase table. This allows you to master multiple large data sources into a single output using PME for BigInsights 11.3.0.
Improved Performance and Algorithm Tuning
Searches can now specify that only specific input sources should be considered when performing probabilistic matching. This new source filtering ability will allow for quicker searches by focusing only on the data of interest. PME for BigInsights 11.3.0 also now supports MDM algorithm features for bucketing roles. Size limits for bucketing roles and settings limiting bucket usage for searching and linking are now honored.