基于混合专家的共享账户推荐算法

Shared-account recommendation with mixture-of-experts

摘要

针对共享账户电视推荐场景下的行为序列混合、主导偏好时序变化问题,提出一种基于混合专家的共享账户推荐算法(MoE-SAR)。该方法通过混合专家网络分解行为序列,并利用动态门控机制自适应融合专家输出,以精准识别个体用户特征。同时,MoE-SAR方法引入对比学习策略,以最小化同源专家,增强样本间距离、最大化不同专家的输出间互信息,有效提升表征区分性与稳定性。此外,该方法还结合Transformer进行时序建模,以精准捕捉共享账户中个体用户的个性化偏好。实验结果表明,在HVIDEO数据集的E域上,MoE-SAR在MRR@20上较次优基线提升了24.0%,在Recall@20上提升了10.8%。在V域上,MRR@20提升了8.0%,Recall@20提升了4.7%。

Abstract

This paper addresses the issues of behavior sequence mixing and temporal changes in dominant preferences in the shared account TV recommendation scenario by proposing a mixture-of-experts shared account recommendation (MoE-SAR) algorithm. This method decomposes behavior sequences using a mixture of experts network and adaptively fuses expert outputs through a dynamic gating mechanism to accurately identify individual user characteristics. Additionally, the MoE-SAR method introduces a contrastive learning strategy to minimize the distance between samples from the same source expert and maximize the mutual information between outputs from different experts, effectively enhancing the discriminability and stability of representations. Furthermore, this approach integrates Transformers for temporal modeling to accurately capture the personalized preferences of individual users within shared accounts. The experimental results indicate that on the E-domain of the HVIDEO dataset, MoE-SAR improves MRR@20 by 24.0% and Recall@20 by 10.8% over the second-best baseline. On the V-domain, MRR@20 improves by 8.0% and Recall@20 improves by 4.7%.

关键词

共享账户 / 混合专家 / 推荐系统

Key words

shared account / mixture-of-experts / recommendation system

参考文献

[1] QIN J R, ZHU J C, LIU Y K, et al. Learning to distinguish multi—user coupling behaviors for TV recommendation[C]// Proceedings of the Sixteenth ACM International Conference on Web Search and Data Mining. Singapore: ACM, 2023: 204-212.

[2] GUO L, YIN H, WANG Q Y, et al. Streaming session—based recommendation[C]// Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining. New York: ACM, 2019: 1569-1577.

[3] 郭磊, 李秋菊, 刘方爱, 等. 基于自注意力网络的共享账户跨域序列推荐[J]. 计算机研究与发展, 2021, 58(11): 2524-2537.

[4] GUO Lei, LI Qiuju, LIU Fang'ai, et al. Shared—account cross—domain sequential recommendation with self—attention network[J]. Journal of Computer Research and Development, 2021, 58(11): 2524-2537.

[5] 高玉凯, 王新华, 郭磊, 等. 一种基于协同矩阵分解的用户冷启动推荐算法[J]. 计算机研究与发展, 2017, 54(8): 1813-1823.

[6] GAO Yukai, WANG Xinhua, GUO Lei, et al. Learning to recommend with collaborative matrix factorization for new users[J]. Journal of Computer Research and Development, 2017, 54(8): 1813-1823.

[7] RENDLE S, FREUDENTHALER C, GANTNER Z, et al. BPR: bayesian personalized ranking from implicit feedback[EB/OL]. (2012—05—09)[2026—01—16]. https://arxiv.org/abs/1205.2618.

[8] MA M Y, REN P J, LIN Y J, et al. π—net: a parallel information—sharing network for shared—account cross—domain sequential recommendations[C]// Proceedings of the 42nd International ACM SIGIR Conference on Research and Development in Information Retrieval. Paris: ACM, 2019: 685-694.

[9] TANG L, GUO L, CHEN T, et al. DA—GCN: A domain—aware attentive graph convolution network for shared—account cross—domain sequential recommendation[C]// Proceedings of the 30th International Joint Conference on Artificial Intelligence. Montreal: IJCAI, 2021: 2483-2489.

[10] WEN X Y, PENG Z H, HUANG S S, et al. MISS: a multi—user identification network for shared—account session—aware recommendation[C]// Database Systems for Advanced Applications. Cham: Springer, 2021: 228-243.

[11] ZHANG C, LIU D Y, ZUO L, et al. Multi—gate mixture—of—contrastive—experts with graph—based gating mechanism for TV recommendation[C]// Proceedings of the 32nd ACM International Conference on Information and Knowledge Management. New York: ACM, 2023: 4938-4944.

[12] BAJAJ P, SHEKHAR S. Experience individualization on online TV platforms through persona—based account decomposition[C]// Proceedings of the 24th ACM International Conference on Multimedia. Amsterdam: ACM, 2016: 252-256.

[13] JIANG J Y, LI C T, CHEN Y, et al. Identifying users behind shared accounts in online streaming services[C]// Proceedings of the 41st International ACM SIGIR Conference on Research & Development in Information Retrieval. New York: ACM, 2018: 65-74.

[14] YANG S, SARKHEL S, MITRA S, et al. Personalized video recommendations for shared accounts[C]// 2017 IEEE International Symposium on Multimedia. Piscataway: IEEE, 2018: 256-259.

[15] WANG Z J, YANG Y, HE L, et al. User identification within a shared account: improving IP—TV recommender performance[C]// Advances in Databases and Information Systems. Cham: Springer, 2014: 219-233.

[16] YANG Y, HU Q M, HE L, et al. Adaptive temporal model for IPTV recommendation[C]// Web—Age Information Management. Cham: Springer, 2015: 260-271.

[17] SUN W C, MA M Y, REN P J, et al. Parallel split—join networks for shared account cross—domain sequential recommendations[J]. IEEE Transactions on Knowledge and Data Engineering, 2023, 35(4): 4106-4123.

[18] GUO L, ZHANG J Y, TANG L, et al. Time interval—enhanced graph neural network for shared—account cross—domain sequential recommendation[J]. IEEE Transactions on Neural Networks and Learning Systems, 2023, 35(3): 4002-4016.

[19] LI D C, LI X, WANG J, et al. Video recommendation with multi—gate mixture of experts soft actor critic[C]// Proceedings of the 43rd International ACM SIGIR Conference on Research and Development in Information Retrieval. New York: ACM, 2020: 1553-1556.

[20] QIN Z, CHENG Y, ZHAO Z, et al. Multitask mixture of sequential experts for user activity streams[C]// Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining. New York: ACM, 2020: 3083-3091.

[21] MA J, ZHAO Z, YI X, et al. Modeling task relationships in multi—task learning with multi—gate mixture—of—experts[C]// Proceedings of the 24th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining. New York: ACM, 2018: 1930-1939.

[22] MA J Q, ZHAO Z, CHEN J L, et al. SNR: sub—network routing for flexible parameter sharing in multi—task learning[J]. Proceedings of the AAAI Conference on Artificial Intelligence, 2019, 33(1): 216-223.

[23] TANG H Y, LIU J N, ZHAO M, et al. Progressive layered extraction (PLE): a novel multi—task learning (MTL) model for personalized recommendations[C]// Proceedings of the 14th ACM Conference on Recommender Systems. New York: ACM, 2020: 269-278.

[24] ZHANG Z, LIU S, YU J, et al. M3oE: multi—domain multi—task mixture—of—experts recommendation framework[C]// Proceedings of the 47th International ACM SIGIR Conference on Research and Development in Information Retrieval. New York: ACM, 2024: 893-902.

[25] ZHANG S, CHEN L, SHEN D, et al. Hierarchical time—aware mixture of experts for multi—modal sequential recommendation[C]// Proceedings of the ACM Web Conference 2025. New York: ACM, 2025: 3672-3682.

[26] SHAZEER N, MIRHOSEINI A, MAZIARZ K, et al. Outrageously large neural networks: the sparsely—gated mixture—of—experts layer[C]// Proceedings of the International Conference on Learning Representations. Toulon: ICLR, 2017.

[27] FEDUS W, ZOPH B, SHAZEER N. Switch transformers: scaling to trillion parameter models with simple and efficient sparsity[J]. Journal of Machine Learning Research, 2022, 23(120): 1-39.

[28] BIAN S Q, PAN X Y, ZHAO W X, et al. Multi—modal mixture of experts representation learning for sequential recommendation[C]// Proceedings of the 32nd ACM International Conference on Information and Knowledge Management. Birmingham: ACM, 2023: 110-119.

[29] HE X N, LIAO L Z, ZHANG H W, et al. Neural collaborative filtering[C]// Proceedings of the 26th International Conference on World Wide Web. New York: ACM, 2017: 173-182.

[30] HE X, DENG K, WANG X, et al. LightGCN: simplifying and powering graph convolution network for recommendation[C]// Proceedings of the 43rd International ACM SIGIR Conference on Research and Development in Information Retrieval. New York: ACM, 2020: 639-648.

[31] CAI X H, HUANG C, XIA L H, et al. LightGCL: simple yet effective graph contrastive learning for recommendation[C]// Proceedings of the Eleventh International Conference on Learning Representations. Kigali: ICLR, 2023: 1-15.

[32] HU G, ZHANG Y, YANG Q. CoNet: collaborative cross networks for cross—domain recommendation[C]// Proceedings of the 27th ACM International Conference on Information and Knowledge Management. Torino: ACM, 2018: 667-676.

[33] HIDASI B, KARATZOGLOU A, BALTRUNAS L, et al. Session—based recommendations with recurrent neural networks[EB/OL]. (2015—10—21)[2026—01—15]. https://arxiv.org/abs/1511.06939.

[34] ZHU Y, LI H, LIAO Y, et al. What to do next: modeling user behaviors by time—LSTM[C]// Proceedings of the International Joint Conference on Artificial Intelligence. Melbourne: ACM, 2017: 3602-3608.

[35] ZHAO W X, MU S, HOU Y, et al. RecBole: towards a unified, comprehensive and efficient framework for recommendation algorithms[C]// Proceedings of the 30th ACM International Conference on Information & Knowledge Management. Queensland: ACM, 2021: 4653-4664.

[36] ZHAO W X, HOU Y, PAN X, et al. RecBole 2.0: towards a more up—to—date recommendation library[C]// Proceedings of the 31st ACM International Conference on Information & Knowledge Management. New York: ACM, 2022: 4722-4726.

[37] WANG X H, YUE H P, GUO L, et al. User identification network with contrastive clustering for shared—account recommendation[J]. Information Processing & Management, 2025, 62(3): 104055.

[38] GUO L, ZHANG J Y, CHEN T, et al. Reinforcement learning—enhanced shared—account cross—domain sequential recommendation[J]. IEEE Transactions on Knowledge and Data Engineering, 2023, 35(7): 7397-7411.

[39] GUO L, WANG C X, WANG X H, et al. Automated prompting for non—overlapping cross—domain sequential recommendation[J]. IEEE Transactions on Knowledge and Data Engineering, 2025, 37(9): 4990-5003.

基金资助

国家自然科学基金资助项目(62372277)