Weka is a collection of machine learning algorithms for data mining tasks. The
algorithms can either be applied directly to a dataset or called from your own
Java code. Weka contains tools for data pre-processing, classification,
regression, clustering, association rules, and visualization. It is also
well-suited for developing new machine learning schemes.

WWW:	http://www.cs.waikato.ac.nz/ml/weka/