A Comparative Study of Collaborative Movie Recommendation System

Abstract

The number of movies available has expanded, making it challenging to select a film that uses current technology to meet users' needs. Following the widespread use of internet services, recommendation systems have become commonplace. The objective for all recommendation systems now is to employ filtering and clustering algorithms to recommend content users are interested in. Suggestions for a media commodity like movies are offered to consumers by locating user profiles of people with comparable likes which makes users' preferences initially determined to allow them to rate movies of their choosing. After a period of use, the recommender system understands the user and offers films that are more likely to receive higher ratings. A comparison study on the existing models helps to understand future scope and improvements for more personalized models for movie recommendation. In comparison to previous models, the MovieLens dataset gives a dependable model that is exact and delivers more customized movie suggestions. In this paper, an approach to do a detailed study and review the user preferences based on item and content of movies has been made to understand the filtering techniques of the collaborative recommendation system to increase accuracy and give highly rated movies as recommendations to the user is carried and based on the results the recommendation system is built with a content-based filtering technique.