REACTION: A naive machine learning approach for sentiment classification
Silvio Moreira, João Filgueiras, Bruno Martins, Francisco M. Couto, Mário J. Silva · Joint Conference on Lexical and Computational Semantics · 2013
We evaluate a naive machine learning approach to sentiment classification focused on Twitter in the context of the sentiment analysis task of SemEval-2013. We employ a classifier based on the Random Forests algorithm to determine whether a tweet expresses overall positive, negative or neutral sentiment. The classifier was trained only with the provided dataset and uses as main features word vectors and lexicon word counts. Our average F-score for all three classes on the Twitter evaluation dataset was 51.55%. The average F-score of both positive and negative classes was 45.01%. For the optional SMS evaluation dataset our overall average F-score was 58.82%. The average between positive and negative Fscores was 50.11%.