Movie Recommendation System Using Hybrid Collaborative Filtering Model

Recommendation systems play an important role in e-commerce marketing and online streaming, such as Netflix, YouTube, and Amazon (Chin et al. in A fast parallel stochastic gradient method for matrix factorization in shared memory systems, 2015 [1]). Making the right recommendation for the next product, music, or movie increases user retention and satisfaction, leading to sales and revenue growth. The hybrid model is a combination of user-based collaborative filtering and item-based collaborative filtering and is using the cosine similarity and jacard index (Chin et al. in A Learning-rate Schedule for Stochastic Gradient Methods to Matrix Factorization, 2015 [2]) to find the best results possible from the MovieLens dataset that the model is using. The system is combination of random prediction, regularization, linear model, and matrix factorization to get the best plausible results (Chin et al. in LIBMF: a library for parallel matrix factorization in shared-memory systems, 2016 [3], Subramaniyaswamy in A personalized movie recommendation system based on collaborative filtering [4]).