Enhancing Movie Recommendations: A Deep Neural Network Approach with MovieLens Case Study

Abstract

Movie recommendation systems are a hot topic in research these days, thanks to the easy access to fast internet and the growing love for movies and multimedia entertainment. But it’s not just about suggesting what movie to watch next; these systems are also handy for suggesting courses or products when you’re shopping online. They’re even useful for spreading information around. Building recommendation systems is challenging due to issues like the cold start problem, sparsity of data, the long-tail problem, and the lack of explicit feedback. This research emphasizes the need for advanced recommendation systems to navigate the abundance of choices we encounter daily, highlighting the importance of making strategic recommendations. Our research delves into the realm of Deep Learning (DL), employing sophisticated methodologies such as Deep Neural Networks (DNNs), Convolutional Neural Networks (CNNs), Recurrent Neural Networks (RNNs), and AutoEncoders (AEs). While previous studies have explored these techniques, our primary emphasis lies in enhancing deep collaborative neural networks to augment recommendation systems reliant on implicit feedback. Our approach aims to enhance recommendation system performance by addressing the limitations of traditional algorithms, particularly in scenarios where explicit user feedback is lacking. We evaluated our method using a variety of assessment metrics, incorporating loss functions, success rates (Hit @ ratio), and relevance scores (NDCG @ratio). Excitingly, our experiments demonstrate that our approach surpasses prior methods in recommending items based on indirect feedback. This represents a significant advancement in the development of more intelligent and efficient recommendation systems.