Principal Component Analysis based Opinion Classification for Sentiment Analysis

Sentiments express perspectives or opinions of users, and reviews gives information about how a product is seen. Online reviews are famously utilized for judging quality of product or service and impact decision making of the users while selecting a product or service. Sentiments are progressively accessible in type of reviews and feedback at websites, blogs, and micro blogs, which impacts future customers. As it is not doable to physically handle the colossal measure of sentiments created online, Sentiment Analysis utilizes automatic processes for extracting reviews and separate significant information with sentiment orientation. In this paper, it is proposed to extract the feature set from movie reviews. Reverse document frequency is computed and the feature set is reduced utilizing Principal Component Analysis. The effectiveness of the pre-processing is evaluated utilizing Naive Bayes and Linear Vector Quantization.