An Enhanced Feature Selection for Text Documents

In the current digital world, a vast amount of data is recorded in a variety of forms like pictures, data, video, and audio. Generally, such type of information which is available in large voluminous form is actually not available in an organized manner which is appropriate for text processing. Text mining is a subfield of data mining which aims at exploring the useful information from the recorded resources. Document clustering helps the users to effectively navigate, review, and classify text documents into significant clusters, the knowledge that helps to handle the enormous amount of text mining. Preprocessing and feature selection are of tremendous importance in document clustering. In document clustering, preprocessing techniques applied to the documents are Bag of Words (BOW), Stop word removal, and Porter stemming. In this paper, we proposed an easy to use framework for preprocessing and Enhanced Term Frequency-Inverse Document Frequency (Enhanced TF-IDF) method for feature selection.