Neural Embedding-Based Collaborative Filtering for Movie Recommendation Services

This study introduces a cutting-edge approach, Movie2Vec, that uses neural embedding for collaborative filtering to enhance user experience. Leveraging the power of neural networks, Movie2Vec employs innovative item embedding techniques to capture intricate relationships between movies and user preferences. By delving into collaborative filtering, the model not only refines its understanding of individual user tastes but also taps into the collective wisdom of a diverse user base. The manuscript proposes and describes the Movie2Vec framework, highlighting its architecture, training process, and performance evaluation. The experiments implemented on the MovieLens datasets (25M and Lastest) demonstrate the efficacy of the proposed method in surpassing traditional collaborative filtering-based recommendation methods in terms of Precision, Recall, and F1 evaluation metrics.