Movie Recommendations with Conventional Strategies using MovieLens Dataset

Abstract

World Wide Web is extending in an exponential rate, the size and the complexity of the real-world data is expanding alongside it. Currently, the web contains Terra bytes of data and most of the part doesn’t interest by the users or either undesirable data or substance irrelevant information to user’s choice. To enable the users to cope with this data explosion, many organizations deploying tools of recommendation systems to guide the users in the right direction and to get the benefit themselves in terms of the business growth. Many e-commerce systems using these recommendation systems to recommend books, articles, news, products, and movies, etc. Among these, Movie recommendation systems have turned into an intriguing research area, because of the exponential increase of the users in a mobile environment. For such kind of movie recommendations, the aggregated data which include preferences of the users, feelings of the users, and reviews by the users take a key role to assist the new users for taking advantage. In this paper we presented a brief overview of the recommender systems using content-based filtering (preference of the user), collaborative filtering (preference of similar users), and hybrid-based filtering. We applied these three strategies to the minimized sample from the MovieLens dataset. However to deal with the recommendation system, we must consider timeliness and accuracy. In addition we have presented the recent works on movie recommendations using machine learning strategy and directions to be taken for getting timely and accurate recommendations