Filtering trolling comments through collective classification

Abstract

Nowadays, users are increasing their participation in the Internet and, particularly, in social news websites. In these webs, users can comment diverse stories or other users' comments. In this paper we propose a new method based for filtering trolling comments. To this end, we extract several features from the text of the comments, specifically, we use a combination of statistical, syntactic and opinion features. These features are used to train several machine learning techniques. Since the number of comments is very high and the process of labelling tedious, we use a collective learning approach to reduce the labelling efforts of classic supervised approaches. We validate our approach with data from 'Menéame', a popular Spanish social news site. © 2013 Springer-Verlag.