WEKA

Found only on the islands of New Zealand, the Weka is an endangered flightless bird with an inquisitive nature. So it is only fitting that the only state-of-the-art Data Mining suite, developed for almost a decade at the University of New Zealand, is also called WEKA.

I have begun using WEKA somewhere in 2000/2001 at the beginning of my PhD. My topic was ensemble learning, so I was searching for not only an integrated set of machine learning algorithms, but also a suite where it was easy to combine, mix/match and add new scheme. Oh, and it should be state-of-the-art, and of course Open Source. WEKA fit all these requirements and more, and I have been a major champion and minor contributor of code as well as observant reader of the WEKA mailing list ever since - and still use it for almost all of my machine learning and data mining needs.

It is a little known fact that one of European's most popular Data Mining Suite, RapidMiner, includes WEKA as a major part of their Open Source version. But see for yourself. Sadly, the company does not adequately credit WEKA and its role in their continuing success. This has been amply demonstrated by our discussion at KDnuggets.

However, the main reason for this page is that I am the original author of the command-line guide to WEKA, which has been put in a Wiki for everyone to improve upon. Here it is - have fun!