Weka is developed by the University of Waikato in New Zealand. The project, which was started in 1993 and available on sourceforge.net since 2000, was originally designed to provide a platform for the research and testing of advanced machine learning techniques.
It has since evolved to include advanced analytic and data mining algorithms and visualization capabilities to identify complex patterns, associations and correlations in data, making it highly applicable for evaluating performance and supporting strategic planning initiatives for commercial organizations in retail, financial services, communications and consumer packaged goods.
Weka has been downloaded more than 600,000 times on sourceforge.net and has spawned over two dozen community developed projects.
Mark Hall, a chief developer of Weka at Waikato University, said the time was ripe for Pentaho to pump additional resources into the project. The popularity of Weka has grown to the point where it needs support from a larger open source community.
Following the acquisition, Weka is now also immediately downloadable from Pentaho’s website as a general public license. Pentaho has also launched a live forum for interaction among Weka project community members.
Pentaho said it will announce a road map detailing further enhancements to Weka soon. The plans include tight integration with the rest of the Pentaho BI Suite that also includes OLAP analysis, reporting, workflow, dahsboarding and data integration modules. It also plans to extend the Weka code to additional vertical applications.
Andre Boisvert, chairman and co-founder of Orlando, Florida-based Pentaho, said that Weka was an excellent product with a vibrant worldwide development community behind the project. He should know; he served as the former COO of SAS Institute Inc, which is a leader in data mining software.
Pentaho’s acquisition is the latest in a series of acquisitions by the firm to assemble a comprehensive, integrated BI platform.
In November last year Pentaho acquired the Java-based Mondrian OLAP Server project, making it a permanent fixture in its BI platform.
The company followed that up earlier this year by adding enterprise reporting, through the inclusion of the JFreeReports project and proprietary software acquired from GridVision Engineering GmbH, and more recently data integration, via the April acquisition of the Kettle ETL project.
Pentaho continues to show that there is no capability or application in the BI market that can’t or won’t be addressed by the open source community, Boisvert said.
Pentaho was founded in 2002 by former executives of leading commercial software firms like IBM, Hyperion Solutions, Cognos, Oracle and SAS. The company is effectively stewarding the development of a LAMP-like BI stack that pulls together a set of interoperable open source BI tools into a comprehensive integrated BI stack.
The first general release of the platform was delivered last December.
Other projects associated with Pentaho include JPivot, Firebird RDBMS, Enhyra Shark, and Eclipse BIRT.
Pentaho is part of a growing movement of open source BI upstarts like Jaspersoft Inc, Greenplum Inc and Kinetic Networks Inc that claim to develop less expensive alternatives to premium priced commercial BI technologies from Cognos Inc and Business Objects SA.
