An Effective TF-IDF Model to Improve the Text Classification Performance
Shitanshu Jain, Sapan Kumar Jain, Somil Vasal
Abstract
Term weight is utilized as a baseline classifier with text classification and other text mining techniques used for a significant increase in efficiency. The words, documents, and datasets are taken into consideration while applying various term weighting techniques. In this paper, various TF-IDF models are evaluated for text classification performance, and a proposed TF-IDF is suggested. The TF-IDF model's mathematical definition that we put forward was based on the computation of term frequency, term occurrence, TF-IDF Variants, and typical TF-IDF. Our aim is to test various phrase weighting techniques on the Amazon review dataset. With respect to accuracy performance parameters employing various classifiers, the suggested method outperforms the other weighting system. This A research indicates that modifications to the original TF-IDF weighting techniques offer considerable gains in