Multi-Label News Category Text Classification

Abstract

                News classification in the present day is a challenging research problem due to availability of huge digital data demanding the accurate classification for user easy. The world is witnessing the transformation in digital age giving rise to change in every aspect of human life. Each news category arranges the news story prior to distributing it. So every time guests visit their site, can without much of a time find news of their interests. Presently, the news stories are ordered by hand by the substance chiefs of information sites. The aim of our article is to build a text classifier for news category and further analyze the sentiments of text-based headlines. The Experimental study is carried on 10 News categories available in dataset (‘News_Category_Dataset_v2’ from Kaggle). The results reveal sentiment associated with news categories as polarity index with positive or negative values.