Semantic Annotation of Classification Data for KDD Support Services

In order to guide the user in the correct and effective use of tools, a KDD support system should posses a knowledge on the mapping between certain characteristics of data (and the Data Mining task) and the tools which could be profitably used. In this work, we want to dwell upon this problem in the framework of the classification data mining task, by showing an approach to extract some knowledge about the data from the form of the decision border.