Bayesware Discoverer, the first Knowledge Discovery program based on Bayesian Networks, will be available for sale in January 2001. Bayesian Networks are one of the most successful results of recent Artificial Intelligence research and Bayesware Discoverer is the first program to apply this technology to Knowledge Discovery and Intelligent Data Analysis.

Bayesware Discoverer is the industrial strength version of Bayesian Knowledge Discoverer (BKD), a research prototype distributed in over 3,000 copies worldwide. After two years of prototype testing, Bayesware Limited has taken up the development of a commercial version.

Bayesware Discoverer is the result of these testing and development efforts. Bayesware Discoverer is over one order of magnitude faster than the original prototype and includes some of the most demanded features identified during the prototype testing phase: a more sophisticated unsupervised discretization able to handle an arbitrary number of states using both median and mean values, and the ability to predefine unchangeable dependencies. These features make Bayesware Discoverer the ideal tool not only for learning Bayesian Networks but also for Bayesian supervised classification and feature dependency selection.

For further information, read the description of Bayesware Discoverer.