Research on movie recommendation algorithm based on graph neural network

In this paper, we present an innovative approach to enhancing the efficacy and personalization of movie recommendation systems by incorporating a Graph Neural Network (GNN) model augmented with an attention mechanism. This model leverages the intricate structure of user-movie interactions, encapsulated within a graph framework, to decipher the nuanced interplay between user inclinations and cinematic attributes. It further elucidates how these preferences and characteristics disseminate through the network, affecting recommendations in a holistic manner. The integration of the attention mechanism is pivotal, enabling the model to dynamically ascertain the varying degrees of importance that distinct users place on specific movie features. This feature-wise weighting allows for a granular adjustment of recommendation outcomes, tailoring them precisely to individual tastes. Essentially, the attention mechanism empowers the GNN to focus on the most salient aspects of each movie for each user, thereby refining the recommendation process to unprecedented levels of personalization. Our empirical evaluations substantiate the superiority of this novel model over conventional recommendation algorithms. Across a spectrum of performance metrics, our GNN-based system demonstrates marked enhancements. This translates into not only more accurate predictions but also a significantly enhanced user experience, characterized by recommendations that resonate closely with personal preferences.