Tackling the poor assumptions of naive bayes text classifiers

Abstract

Naive Bayes is often used as a baseline in text classication because it is fast and easy to implement. Its severe assumptions make such eciency possible but also adversely af-fect the quality of its results. In this paper we propose simple, heuristic solutions to some of the problems with Naive Bayes classiers, ad-dressing both systemic issues as well as prob-lems that arise because text is not actually generated according to a multinomial model. We nd that our simple corrections result in a fast algorithm that is competitive with state-of-the-art text classication algorithms such as the Support Vector Machine. 1.