Multi-Label Text Classification using Machine Learning Techniques

Abstract

In day-to-day life, the data growth rate in internet surges and handling the data becomes a
complex task. Classifying the unstructured data and retrieving the useful content plays a
major part in digital technologies. As lot of news generated daily from different sources on
the internet there is always a need to categorize the news articles in order to provide valuable
information to the news readers on time. The paper focus to build multilabel text
classification for AG News dataset using various classification algorithms. It also discusses
the steps involved in news articles classification and implement Logistic Regression, Naïve
Bayes, Support Vector Machine and Multi-Layer Perceptron classification algorithms on AG
News dataset containing articles on four different categories. The investigations revealed that
when compared to other three classifiers, performance of naïve bayes is superior, and the NB
achieves a high accuracy of 90 percent.