The MiningMart system V0.21b supports some specific implementations of
Machine Learning algorithms using a wrapper approach. Currently the
following tools are supported if put to the directory
	<MMart_directory>/compiler/runtime/bin/<platform>/ :

* mySVM, a Support Vector Machine by Stefan Rueping
For details please refer to
	http://www-ai.cs.uni-dortmund.de/SOFTWARE/MYSVM/
The wrapper expectes the target files "mysvm" and "predict".

* Apriori, implemetation of Bart Goethals
For details please refer to
	http://www.cs.helsinki.fi/u/goethals/software/index.html
The wrapper expects the binary "apriori".

* C4.5 by Quinlan (SunOS only):
The software may be downloaded at
	http://www.cse.unsw.edu.au/~quinlan/
Expected binaries are "c4.5" and "c4.5rules"

On SunOS additional operators are available. Please download the file
MMart_SunOS_plugins.tar.gz from the MiningMart download page.

To enable a smooth integration of the 'YALE' learning environment with
MiningMart preprocessing chains it is possible to apply YALE models,
similar to the operator 'ModelApplier' of YALE directly to database
views in MiningMart. Please download YALE (version 2.1 or newer) from
	http://yale.cs.uni-dortmund.de/ ,

install it and make sure that the JAR-files 'yale.jar' and 'weka.jar'
found in YALE's lib directory are in the CLASSPATH when starting
MiningMart. Please refer to the UserGuide on how to apply YALE models
in MiningMart.

