An application of intuitionistic fuzzy sets in text classification

Intuitionistic fuzzy set (IFS) is an extended version of fuzzy set being capable of representing hesitancy degrees. Based on similarity measures for IFSs, a framework for text categorization is presented. Two main challenges are addressed: one is how to represent documents in terms of IFSs; the other is how to learn a pattern of each category from such IFS-based representation. As an exploratory study, the proposed framework is applied to a benchmark data set for text categorization. By using some existing similarity measures for IFSs, the experimental results show that the proposed framework yields satisfactory results.