Recommendation System using Lexicon Based Sentimental Analysis with collaborative filtering

Abstract

Looking at the current scenario it’s easy to come to the conclusion that it’s all about data. Data is being generated at an exponential rate. Not only the amount of data being generated is exponential but also the rate and variety are also accelerating. This also includes a tremendous increase in growth of service, customers and online reviews. More than trillion of reviews of various products are found on the internet. Earlier recommendation systems often put up efficiency problems. In this paper, we present the recommendation system to the customers based on their interest. Sentimental analysis is used to recommend based on the top k rated reviews based on top positive reviews of a geographically wise. Here we use the Hadoop Framework to make our system more efficient and scalable.