Performance of Hybrid Sentiment Classification Model on Online Product Reviews

Sentiment classification attempts to identify the sentiment polarity of a given text as either positive or negative. Much of the work has been focused on Sentiment classification using machine learning methods in last decades. Analyzing and predicting the polarity of the sentiment plays an important role in decision making. Related work about hybrid methods contributing to sentiment classification are still limited and more extensive experimental work is needed in this area. In this study sentiment classification is done using hybrid method with support vector machine (SVM) as base classifier. The results show that hybrid model performs better in terms of error rate and receiver operating characteristics curve (ROC) for various sampling methods.