A Movie Recommendation System:Using Content-based and Collaborative Information

Abstract

Since the late 20th century, the number of internet users has increased dramatically as has the number of web searches performed on a daily basis and the amount of information available to us. However, not all data that we get in search results are reliable or relevant which means that it may become more and more difficult to get satisfactory results from web searches. To solve this problem we use recommendation systems. Recommendation system shows us only the relevant information. Recommendation Engines can make various suggestions about artifacts to users. In our day-to-day lives, they may predict whether a user may like to buy a particular product online or if they are interested in a particular movie or are interested in listening to a particular song. To pick a movie, a user might search various websites to find a highly rated and well-reviewed movie which is very time-consuming. Here, we have created this recommendation system using the following methods. Content-based filtering takes keywords from the movie dataset and suggests it to a relevant user. Social based filtering takes reviews from multiple users and suggests it to another user. However, these methods do not use a significant amount of information available. This paper is an approach to a recommendation that is able to use both user ratings and other information available that will help in recommending movies to users. Our method uses these methods on a dataset containing more than 5000 different movies.

Index Terms:Movie recommendation system, Collaborative filtering, content-based filtering, Hybrid approach, Scalability