Accelerating Generative Recommendation via Simple Categorical User Sequence Compression

Authors: Q Liu, L Fan, Z Liu, X Dong, Y Luo, G An, N Chen, W Guo, Y Liu, XM Wu
Time: 2026
Journal: Proceedings of the Nineteenth ACM International Conference on Web Search and Data Mining (WSDM)

Abstract

Although generative recommenders demonstrate improved performance with longer sequences, their real-time deployment is hindered by substantial computational costs. To address this challenge, we propose a simple yet effective method for compressing long-term user histories by leveraging inherent item categorical features, thereby preserving user interests while enhancing efficiency. Experiments on two large-scale datasets demonstrate that, compared to the influential HSTU model, our approach achieves up to a 6x reduction in computational cost and up to 39% higher accuracy at comparable cost (i.e., similar sequence length).

Keywords: generative recommendation, user sequence compression, efficiency

Citation

Q Liu, L Fan, Z Liu, X Dong, Y Luo, G An, N Chen, W Guo, Y Liu, XM Wu, "Accelerating Generative Recommendation via Simple Categorical User Sequence Compression," Proceedings of the Nineteenth ACM International Conference on Web Search and Data Mining (WSDM), 2026.