A Fuzzy Approach to Classification of Text Documents

This paper discusses theclassification problems of text documents. Based on the concept of theproximity degree, the set of words is partitioned into some equivalenceclasses. Particularly, the concepts of the semantic field andassociation degree are given in this paper. Based on the above concepts,this paper presents a fuzzy classification approach for documentcategorization. Furthermore, applying the concept of the entropy ofinformation, the approaches to select key words from the set of wordscovering the classification of documents and to construct thehierarchical structure of key words are obtained.