Smart Recommendation System Based on Product Reviews Using Random Forest

Abstract

Social network, e-commerce sites, blogs are new emerging platforms for people to express their opinion. These sites contain huge amount of text which can be used for different purpose like Sentiment Analysis. Sentiment Analysis is a growing field in natural language processing. Sentiment analysis is major focused on company's improvement. But sentiment analysis can be useful in recommendation system also. Based on various performance measures, this paper compares the results of machine learning algorithms like Multinomial Naive Bayes algorithm, Logistic Regression, SVM Classifier, Decision Tree and Random Forest. These algorithms are used for sentiment analysis of reviews and in turn for product recommendation. In proposed system, Random Forest shows outstanding performance. To create suitable recommendations using the analysis of emotions, there is a need to use polarity obtained through the reviews.