EMOTION CLASSIFICATION OF NEWS HEADLINES USING SVM

Abstract

There is an increasing body of research on understanding the human emotion. In this paper we propose a system for automatic classification of reader’s emotions in anger, disgust, fear, jay, sadness and surprise on the Word Net Affect dataset. For the classification we have used Support Vector Machines with a total of 1000 news headlines provided by “Affective Task” in Sem Eval 2007 workshop which focouses on classification of emotions in text. We have compared our results with those obtained by three systems participating in the SEMEVAL emotion annotation task: SWAT, UPAR7 and UA. Our experiments showed that SVM classification gives better performance for emotion detection in sentences.