New fuzzy decision tree model for text classification

Abstract

In this paper, a supervised automatic text documents classification using the fuzzy decision trees technique is proposed. Whatever the algorithm used in the fuzzy decision trees, there must be a criterion for the choice of discriminating attribute at the nodes to partition. For fuzzy decision trees two heuristics are usually used to select the discriminating attribute at the node to partition. In the field of text documents classification there is a heuristic that has not yet been tested. This paper tested this heuristic.