Litcius/Paper detail

Towards a Personalized Movie Recommendation System: A Deep Learning Approach

Zhongtai Qin, Mingjun Zhang

202117 citationsDOI

Abstract

In view of the limitations of existing personalized movie recommendation algorithms, in this article, we propose a personalized movie recommendation system based on deep neural networks. The system uses a deep neural network to process discrete features to fully dig out the various features between the user and the movie, and construct a user feature vector model and a movie feature vector model with demographic characteristics. Specifically, we introduce an attention mechanism into the model to learn the user's behavioral preferences and integrate user feature vectors to obtain the user's dynamic short-term interest features and static long-term historical interest features. This method enhances the recommendation performance of the model. We conduct performance exploration on the public dataset MovieLens, and the proposed recommendation model is superior to traditional recommendation algorithms in terms of accuracy and recall, and has higher recommendation performance.

Topics & Concepts

MovieLensComputer scienceRecommender systemConstruct (python library)Feature (linguistics)Artificial intelligenceDeep learningProcess (computing)Artificial neural networkPrecision and recallFeature vectorMachine learningRecallCollaborative filteringTerm (time)Recurrent neural networkData miningInformation retrievalPhilosophyOperating systemProgramming languagePhysicsQuantum mechanicsLinguisticsGenerative Adversarial Networks and Image SynthesisAdvanced Data and IoT TechnologiesCinema and Media Studies