View article

Sentiment analysis is treated as a classification task as it classifies the orientation of a text into either positive or negative. This paper describes experimental results that applied Support Vector Machine (SVM) on benchmark datasets to train a sentiment classifier. N-grams and different weighting scheme were used to extract the most classical features. It also explores Chi-Square weight features to select informative features for the classification. Experimental analysis reveals that by using Chi-Square feature selection may provide significant improvement on classification accuracy.