Predicting drop-out from social behaviour of students

This paper focuses on predicting drop-out and school failure when student data has been enriched with data derived from students social behaviour. These data describe social dependencies gathered from e-mail and discussion boards conversation, among other sources. We describe an extraction of new features from both student data and behaviour data (or more precisely from social graph which we construct). Then we introduce a novel method for learning classier for student failure prediction that employs cost-sensitive learning to lower the number of incorrectly classified unsuccessful students. We show that a use of social behaviour data results in significant prediction accuracy increase.