With the tremendous growth in the Internet technologies and E commerce platforms, many people are showing interest in procuring products and services online. in this scenario recommender systems are a great way to promote new business and get more customers for the relevant products and servicesAnd thereby increase the business. However, it is crucial to be able to predict what a customer would like andneed based on the products he has purchased or he is interested in. In order to achieve this the machine learning model would have to accomplish various tasks such as word segmentation, stop-words, extraction of features and finding similar products other users have purchased etc. In this project we take the example of movie recommendation system and we tried to categorize the movie reviews as positive or negative using sentiment analysis and have built a recommender system using an improved item based collaborative filtering based on the sentiment of users which can suggest movies that a user may like based on the list of movies he has already watched.