A Survey on News Classification Techniques

With the advancement and decentralization over the past two decades, the internet has become the most democratic of all mass media. Online news organizations have emerged during this period making news accessible to people which has become efficient, faster and convenient for the readers. The development of information technology and these organizations have led to an increase in the amount of disarrayed data, i.e. news, available over the internet. Over the past years researchers have come up with many techniques to achieve news classification, but the pursuit for higher precision of these techniques is incessant. The modern approaches by researchers to classify news items make use Support Vector Machines (SVMs), Naïve Bayes algorithm, Hierarchical Multi-label Classification, Bayesian Network, Clustering algorithms, etc. In this paper we have studied various research works and techniques and examined their robustness and accuracy to classify news items.