Association rules of fuzzy soft set based classification for text classification problem

Text classification is imperative in order to search for more accessible and appropriate information. It utilized in various fields, including marketing, security, biomedical, etc. Apart from its usefulness, the available classifiers are vulnerable to two major problems, namely long processing time and low accuracy. They can result from a large amount of data presented in the text classification problem. In this paper, we propose a model called Class-Based Fuzzy Soft Associative (CBFSA). This model is a combination of the association rules method and fuzzy soft set model. We used Fuzzy Soft Set Association Rules Mining for generating classifiers and Fuzzy Decision Set of an FP-Soft Set for building classifiers. Our experiment for the 20 Newsgroups dataset on 20 class documents has shown that CBFSA is more accurate than the other soft set classifiers: Soft Set Classifier (SCC), Fuzzy Soft Set Classifier (FSSC) and Hybrid Fuzzy Classifier (HFC). Besides that, it has also shown that CBFSA is more accurate and efficient compared to other associative classifiers such as the classification Based on Association (CBA) method. (C) 2020 The Authors. Production and hosting by Elsevier B.V. on behalf of King Saud University.