Pentaho buys open-source data-mining project

exclusive Pentaho has acquired the Weka open-source data-mining project to fill out its business intelligence product suite.

The acquisition, announced on Tuesday, means the primary five contributors to the Weka project will be employees at Pentaho. Financial terms were not disclosed.

The Weka software, which came from the University of Waikato in New Zealand, is designed to analyze data trends. Retail companies, for example, use data-mining tools to segment customers and devise targeted marketing campaigns.

The software was developed in the mid 1990s and is widely used, said Richard Daley, the CEO of Pentaho.

The addition of the data-mining software creates a full-range of business intelligence tools and allows Pentaho to compete with data-mining specialists SAS and SPSS, Daley said.

Daley said the open-source business model, where Pentaho sells a subscription service to customers and has a free version of its product, allows the company to offer lower prices than established vendors, such as Cognos and Business Objects.

Also, Pentaho engineers were able to design the company's Java-based products from scratch, and that has helped the company compete on features, he said.

"Customers usually bring us in on price first and once they start evaluating, they find that we're technically superior," he said.

The company intends to integrate Pentaho's data-extraction tools with the Weka software, he added. Pentaho has also developed open-source reporting and analytical tools.

"Having the (Weka) people onboard (as employees) helps make sure we're moving in the same direction in terms of a business intelligence suite. It's much easier to administer and be steward under one umbrella," Daley said.

[Savio.Rodrigues]

Umm...can you really buy an open source project

You can buy committers, but not an open source project...or else, it's not a real open source project.

More here: http://saviorodrigues.wordpress.com/2006/09/21/got-community-redux/

Savio

[gashlerm]

Waffles vs. WEKA

WEKA is slow (because it's written in Java), and has license overhead (because it's GPL). The Waffles project is all C++, and it's distributed under the LGPL, which is probably too open to get bought out. If Pentaho manages WEKA to death, open source will still carry on.

[datahelper]

New Data Crunching Programming Language

There is a top notch new data crunching programming language
and data crunching back engine available at:

www.my.opera.com/datahelper

I should remark that what we call "data analysis" can often be
divided into two sub-branches:

data analysis = data crunching + statistical analysis

Statistical analysis is the selection of mathematical algorithms
for data that has already been cleaned up and prepared.

Data crunching, which involves writing much more code and can
be quite time consuming, consists of merging, conforming,
transforming, preparing and otherwise cleaning up data that is
often fairly messy.

Robert ( callingrw@yahoo )