Banu is developed with generating medical knowledge in mind. The program has three main features:

• Automated generation of (adaptable) naive Bayesian networks.
• Connectivity with ILP systems such as Toplog and Aleph to find better adaptable naive Bayesian networks
• Exporting knowledge in Notation3 rules.

Adaptable naive Bayesian networks differ from common naive Bayesian networks by the use of premises. i.e. more specialised, but still naive, Bayesian networks are found, depending on an assumption.

Much more information about the program and the used concepts can be found in this dissertation. The thesis is written in Dutch but there is a summary added in English at the back (under "Bijlage F").

Feel free to contact me with questions or suggestions. ()