SZTE-NLP: Sentiment Detection on Twitter Messages
Viktor Hangya, Gábor Berend, Richárd Farkas · 2013
In this paper we introduce our contribution to the SemEval-2013 Task 2 on “Sentiment Analysis in Twitter”. We participated in “task B”, where the objective was to build models which classify tweets into three classes (positive, negative or neutral) by their contents. To solve this problem we basically followed the supervised learning approach and proposed several domain (i.e. microblog) specific improvements including text preprocessing and feature engineering. Beyond the supervised setting we also introduce some early results employing a huge, automatically annotated tweet dataset. 1