[1]马丽,刘文哲,李雨豪.基于自适应双向信息流的序列推荐[J].郑州大学学报(工学版),2027,48(XX):1-8.[doi:10.13705/j.issn.1671-6833.2026.04.003]
　MA Li,LIU Wenzhe,LI Yuhao.Sequential Recommendation Based on Adaptive Bidirectional Information Flow[J].Journal of Zhengzhou University (Engineering Science),2027,48(XX):1-8.[doi:10.13705/j.issn.1671-6833.2026.04.003]
点击复制

基于自适应双向信息流的序列推荐()
分享到：

《郑州大学学报(工学版)》[ISSN:1671-6833/CN:41-1339/T]

卷:
48
期数:
2027年XX
页码:
1-8
栏目:
出版日期:
2027-12-10

文章信息/Info

Title:
Sequential Recommendation Based on Adaptive Bidirectional Information Flow
作者:
马丽1; 2 ;  刘文哲1 ;  李雨豪1
1. 河北地质大学 信息工程学院,河北 石家庄 052161;2. 河北地质大学 智能传感物联网技术河北省工程研究中心,河北 石家庄 052161
Author(s):
MA Li1; 2 ;  LIU Wenzhe1 ;  LI Yuhao1
1 . College of Information Engineering, Hebei GEO University, Shijiazhuang 052161, China; 2. Intelligent Sensor Network Engineering Research Center of Hebei Province, Hebei GEO University, Shijiazhuang 052161, China
关键词:
序列推荐;  自适应双向信息流;  动态偏好建模;  频域滤波;  注意力机制
Keywords:
sequential recommendation;  adaptive bidirectional information flow;  dynamic preference modeling;  frequency-domain filtering;  attention mechanism
分类号:
TP391. 3TP183F724. 6
DOI:
10.13705/j.issn.1671-6833.2026.04.003
文献标志码:
A
摘要:
通过分析序列推荐中历史与未来信息融合效率欠佳及噪声干扰的问题，提出了一种基于自适应双向信息流的序列推荐方法。在双路径编码器架构基础上，整合了层次化历史总结模块以提炼用户长期偏好，并引入动态频域滤波来抑制数据噪声。所提方法充分考虑了历史与未来信息间的依赖性和交互性，采用自适应双向信息流机制，通过不确定性感知动态调节二者的融合权重，从而精准地刻画用户偏好的演变轨迹。为了验证方法的有效性，在Beauty、Sports、Yelp和ML1M这4个公开数据集上进行了实验，并与10种主流方法进行对比分析。实验结果表明：所提方法在NDCG、HR、MRR等3个指标上均优于对比的基线方法。相较于FMLP-Rec、DualRec与OracleRec这3个头部基线模型，所提方法的HR@20在Beauty和Yelp数据集上分别为0.652 0和0.913 3，比三者的平均水准分别提升了2.08个百分点和2.89个百分点；NDCG@20在Beauty和Yelp数据集上分别达到0.394 4和0.564 5，比三者的平均水准分别提升了2.67个百分点和2.72个百分点。
Abstract:
Addressing the challenges of inefficient information fusion and noise interference in sequential recommendation, a novel method based on an adaptive bidirectional information flow was proposed. Built upon a dual-path encoder architecture, a hierarchical history summarization module was integrated to distill long-term user preferences, and dynamic frequency-domain filtering was introduced to suppress data noise. The approach fully considered the dependency and interactivity between past and future information by employing an adaptive bidirectional information flow mechanism. This mechanism dynamically adjusts fusion weights via uncertainty perception, enabling a precise characterization of the evolution of user preferences. To validate its effectiveness, experiments were conducted on four public datasets including Beauty, Sports, Yelp, and ML1M, and a comparative analysis was performed against 10 mainstream methods. The experimental results demonstrated that the proposed method outperformed the baseline models in three key metrics: NDCG, HR, and MRR. Compared to the three leading baseline models of FMLP-Rec, DualRec, and OracleRec, the proposed method’s HR@20 reached 0.652 0 and 0.913 3 on the Beauty and Yelp datasets, which was 2.08 percentage points and 2.89 percentage points higher than their average performance, respectively. Furthermore, its NDCG@20 on the Beauty and Yelp datasets reached 0.394 4 and 0.564 5, outperforming the average of the three baselines by 2.67 percentage points and 2.72 percentage points, respectively.

参考文献/References:

[1] TANG J X, WANG K. Personalized top-N sequential recommendation via convolutional sequence embedding[C]//Proceedings of the Eleventh ACM International Conference on Web Search and Data Mining. New York: ACM, 2018: 565-573.
[2] HIDASI B, KARATZOGLOU A, BALTRUNAS L, et al. Session-based recommendations with recurrent neural networks[EB/OL]. (2015-11-21)[2025-09-01]. https://arxiv.org/abs/1511.0699.
[3] KANG W C, MCCAULEY J. Self-attentive sequential recommendation[C]//2018 IEEE International Conference on Data Mining (ICDM). Piscataway: IEEE, 2018: 197-206.
[4] SUN F, LIU J, WU J, et al. BERT4Rec: sequential recommendation with bidirectional encoder representations from transformer[C]//Proceedings of the 28th ACM International Conference on Information and Knowledge Management. New York: ACM, 2019: 1441-1450.
[5] WU S, TANG Y Y, ZHU Y Q, et al. Session-based recommendation with graph neural networks[C]//Proceedings of the Thirty-Third AAAI Conference on Artificial Intelligence and Thirty-First Innovative Applications of Artificial Intelligence Conference and Ninth AAAI Symposium on Educational Advances in Artificial Intelligence. New York: ACM, 2019: 346-353.
[6] XU C F, ZHAO P P, LIU Y C, et al. Graph contextualized self-attention network for session-based recommendation[C]//Proceedings of the 28th International Joint Conference on Artificial Intelligence. New York: ACM, 2019: 3940-3946.
[7] ZHOU K, YU H, ZHAO W X, et al. Filter-enhanced MLP is all you need for sequential recommendation[C]//Proceedings of the ACM Web Conference 2022. New York: ACM, 2022: 2388-2399.
[8] YAO Z Y, CHEN X Y, WANG S N, et al. Recommendation transformers with behavior pathaways[C]//Proceedings of the ACM Web Conference 2024. New York: ACM, 2024: 3643-3654.
[9] GAO J T, ZHAO X Y, LI M Y, et al. FMLP4Rec: an efficient JMLP architecture for sequential recommendation systems[J]. ACM Transactions on Information Systems, 2024, 42(3): 1-23.
[10] LIU L M, CAI L, ZHANG C, et al. LinRec: linear attention mechanism for long-term sequential recommender systems[C]//Proceedings of the 46th International ACM SIGIR Conference on Research and Development in Information Retrieval. New York: ACM, 2023: 289-299.
[11] YU J L, XIA X, CHEN T, et al. XSimCLR: towards extremely simple graph contrastive learning for recommendation[J]. IEEE Transactions on Knowledge and Data Engineering, 2024, 36(2): 913-926.
[12] ZHANG H Y, YUAN E M, GUO W, et al. Disentangling past-future modeling in sequential recommendation via dual networks[C]//Proceedings of the 31st ACM International Conference on Information and Knowledge Management. New York: ACM, 2022: 2549-2558.
[13] XIA J F, LI D S, GU H S, et al. Oracle-guided dynamic user preference modeling for sequential recommendation[C]//Proceedings of the Eighteenth ACM International Conference on Web Search and Data Mining. New York: ACM, 2025: 363-372.
[14] SIAMI-NAMINI S, TAVAKOLI N, NAMIN A S. The performance of LSTM and BiLSTM in forecasting time series[C]//2019 IEEE International Conference on Big Data (Big Data). Piscataway: IEEE, 2019: 3285-3292.
[15] WANG W J, FENG F L, HE X N, et al. Denoising implicit feedback for recommendation[C]//Proceedings of the 14th ACM International Conference on Web Search and Data Mining. New York: ACM, 2021: 373-381.
[16] BAI J W, YUAN L, XIA S T, et al. Improving vision transformers by revisiting high-frequency components[C]//Computer Vision-ECCV 2022. Cham: Springer, 2022: 1-18.
[17] MIKHAIL M, OUALI M S, YACOUT S. A data-driven methodology with a nonparametric reliability method for optimal condition-based maintenance strategies[J]. Reliability Engineering and System Safety, 2024, 241: 109668.
[18] XU L, DING F, ZHANG X, et al. Novel parameter estimation method for the systems with colored noises by using the filtering identification idea[J]. Systems & Control Letters, 2024, 186: 105774.
[19] KANG H, YANG M H, RYU J. Interactive multi-head self-attention with linear complexity[EB/OL]. (2024-02-27)[2025-09-01]. https://arxiv.org/abs/2402.17507.
[20] GUO X D, GUO X, LIU Y. SSAN: separable self-attention network for video representation learning[C]//2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Piscataway: IEEE, 2021: 12613-12622.
[21] ZHAO R, WANG K, XIAO Y, et al. Leveraging Monte Carlo dropout for uncertainty quantification in real-time object detection of autonomous vehicles[J]. IEEE Access, 2024, 12: 33834-33399.
[22] HUANG Y X, ZHI X Y, HU J M, et al. FDDA-NET: frequency domain decoupling bidirectional interactive attention network for infrared small target detection[J]. IEEE Transactions on Geoscience and Remote Sensing, 2024, 62: 500416.
[23] MCCAULEY J. Recommender systems and personalization datasets[DB/OL]. [2025-09-01]. http://jmcauley.ucsd.edu/data/amazon/.
[24] Yelp. Yelp open dataset[DB/OL]. [2025-09-01]. https://business.yelp.com/data_resources/open-dataset/.
[25] GroupLens. MovieLens datasets[DB/OL]. [2025-09-01]. https://grouplens.org/datasets/movielens/.

备注/Memo

备注/Memo:
收稿日期:2025-09-20;修订日期:2025-10-15
基金项目:国家自然科学基金资助项目(62476078) ;河北省教育科学规划课题( 2303121) ;河北省高等教育教学改革研究与实践项目(2020GJJG227)
作者简介:马丽(1977— ) ,女,河北张家口人,河北地质大学副教授,博士,主要从事知识发现与粒计算、机器学习的研究,E-mail:mali_new@163.com。
更新日期/Last Update: 2026-02-27