Movie Recommendation System using Cosine Similarity with Sentiment Analysis

Abstract

Multimedia is considered as one of the best sources of entertainment. People of all age groups love to watch movies. Movie Recommender System is essential in our social lives as it enhances the field of entertainment. The proposed system on Movie Recommendation System caters the requirements of the user. The major aim is to provide crisp relevant content to the end-users out of semi-structured content on the internet. The main purpose is to generate accurate, efficient and personalized recommendations to the user. Various building blocks of the paper like Introduction, Literature Survey, Proposed System, Implementation & Result, Comparative Analysis, Conclusion and Future Work are discussed in detail. The proposed machine learning model is trained, tested, and a sentiment classifier is generated which classify the sentiments as a good or a bad sentiment. The recommender system is generated by applying Cosine similarity and making API Calls. As a result, the live working of the system generates accurate and personalized recommendations along with the analysis of sentiments for the end users. It is also concluded that Cosine Similarity provides better and efficient results for a recommender system.