Decision tree algorithms: Integration of domain knowledge for data mining

Abstract

The paper deals with integration of enterprise (domain) knowledge with data mining techniques. The decision tree algorithms CART, ID3, C4.5, CHAID are analyzed in the aspect of relations with enterprise model attributes (i.e. domain knowledge). It is expected, that integration of an enterprise knowledge base in to data mining techniques will improve the data analysis process. The detailed analysis of the modified decision tree algorithm C4.5, integrated with domain knowledge is described. The alignment of business domain attributes used in the decision tree algorithms with the attributes of EM components is defined. © 2012 Springer-Verlag.