Abstract—Integrating knowledge graphs into recommendation systems mitigates data sparsity and cold start issues while offering explanations for recommendations. Capturing user preferences accurately is vital for enhancing performance, but user interests often shift due to contextual and mood changes. Traditional methods overlook preference dynamics, which prompts this paper to propose a knowledge graph based approach considering both long-term and short-term preferences. This approach examines user knowledge graph connections, extracts relevant associations, and integrates them into recommendations. To capture short-term preferences accurately, a bidirectional GRU with an attention mechanism is used. Long-term and short-term preferences are then fused and computed, with optimal parameters determined through experimentation. Experimental results demonstrate that the proposed model improves recall rates by 0.18%, 1.87%, and 1.52% on three datasets, respectively, exhibiting superior performance compared to baseline models.
 
Keywords—knowledge graph, recommendation system, attention mechanism, user preferences


Copyright © 2025 by the authors. This is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited (CC BY 4.0).