Knowledge Graph Based Recommender for Automatic Playlist Continuation

In this work, we present a state-of-the-art solution for automatic playlist continuation through a knowledge graph-based recommender system. By integrating representational learning with graph neural networks and fusing multiple data streams, the system effectively models user behavior, leading to accurate and personalized recommendations. We provide a systematic and thorough comparison of our results with existing solutions and approaches, demonstrating the remarkable potential of graph-based representation in improving recommender systems. Our experiments reveal substantial enhancements over existing approaches, further validating the efficacy of this novel approach. Additionally, through comprehensive evaluation, we highlight the robustness of our solution in handling dynamic user interactions and streaming data scenarios, showcasing its practical viability and promising prospects for next-generation recommender systems.