A Recommendation System for Online Purchase Using Feature and Product Ranking

Social network occupies an important place and takes a considerable amount of time in people's daily life. It has become so popular that people are sharing a huge amount of data and opinion on social network/review sites, which in turn helps to find interesting insights for organizations/vendors or consumers. In this paper, we propose a new algorithm called Feature Based Product Ranking and Recommendation Algorithm (FBPRRA) for providing suggestions to the customers whose are interested in purchasing good quality products. The proposed algorithm analyzes online products and ranks them according to product reviews. Finally, it recommends the suitable product to the target customers. Experiments have been conducted using online reviews for evaluating the proposed algorithm and found that the proposed recommendation algorithm achieves better prediction accuracy than the exiting classifiers such as Naïve Bayes, Support Vector Machine, Random Forest, Decision Tree and K-NN.