An Investigation of Efficient Machine Learning Approaches for Sentiment Classification

Abstract

People’s opinions and experience are very valuable information in decision making process. Now-a-days several websites encourage
users to express their views, suggestions and opinions related to product, services, polices, etc. publically. When a product is purchased by the
customers, the process of quality evaluation is generally takes place. The interpretation of these quality evaluation results or the feelings of the
consumers about the product will be helpful in determining the demand and expectations of the users towards that product. Extracting the useful
content from these opinion sources becomes a challenging task. This paper reviews the machine learning-based approaches to sentiment
analysis and brings out the salient features of techniques in place. The prominently used techniques and methods in machine learning-based
sentiment analysis include - Naïve Bayes, Maximum Entropy and Support Vector Machine, K-nearest neighbour classification.

 

Keywords: opinion; sentiment analysis; machine learning; maximum entropy; support vector machines;