SUPERVISED MACHINE LEARNING ALGORITHMS FOR SENTIMENT ANALYSIS ON IMDB DATA SET

Abstract

In social media, websites, blogs, online product reviews, health care systems and etc. the data
is increasing day by day. Sentiment analysis has been applied on these types of data text is the
largest repository of human based knowledge acquired over thousands of years sentiment
analysis is used to identify, extract and analyze the text data present in the various domains.
In the proposed method seven different types of classifiers are used to predict the dataset
whether it is positive (or) negative among these classifiers. Random forest gives more accurate
performance when compared with the features like precision, Recall and F-Scale in future we
can deploy the unsupervised classifiers to predict the reviews.