Duration-Aware for Short Video Sequential Recommendation

doi:10.3778/j.issn.1002-8331.2401-0030

Abstract

Abstract: Addressing the issues of data sparsity in click data, noise in watch duration feedback, and bias in short video sequential recommendation, a duration-aware for short video sequential recommendation model (DASR) is proposed. This model effectively alleviates the data sparsity issue by deeply modeling user watch duration feedback. Additionally, an unbiased multi-semantic watch duration feedback label generation method is proposed. This method combines the [K]-nearest neighbors algorithm and percentile analysis of training data to dynamically generate label thresholds adapted to different video durations, effectively eliminating the impact of video duration bias. Furthermore, a noise extraction method based on a strong-weak attention network is introduced, accurately extracting positive and negative interest signals from the watch duration, thus addressing the noise issue in watch duration feedback. Extensive experiments on open-source datasets demonstrate that this model outperforms other mainstream methods on multiple evaluation metrics.

Key words: data sparsity, sequential recommendation, attention network, duration bias, dynamic thresholding

摘要： 针对短视频序列推荐中存在的点击数据稀疏性、观看时长反馈中的噪声以及偏差问题，提出了一种基于时长感知的短视频序列推荐模型（duration-aware for short video sequential recommendation，DASR）。该模型通过对用户观看时长反馈的深入建模，有效地缓解了数据稀疏性问题。提出了一种无偏差的多语义观看时长反馈标签生成方法。该方法结合了[K]近邻算法和训练数据的百分位数分析，动态生成适应不同视频时长的标签阈值，有效地消除了视频时长偏差的影响。提出了一种基于强弱注意力网络的噪声提取方法，从观看时长中准确地提取正向和负向兴趣信号，从而解决了观看时长反馈中存在的噪声。在开源的短视频数据集上进行了广泛实验，结果表明该模型在多个评价指标上优于其他主流方法。

关键词: 数据稀疏性, 序列推荐, 注意力网络, 时长偏差, 动态阈值

WANG Hang, YIN Ling, SHI Zhicai, HUANG Bo, GAO Zhirong. Duration-Aware for Short Video Sequential Recommendation[J]. Computer Engineering and Applications, 2024, 60(22): 304-313.

王航, 尹玲, 史志才, 黄勃, 高志荣. 基于时长感知的短视频序列推荐[J]. 计算机工程与应用, 2024, 60(22): 304-313.

References

[1] WANG S, ZHANG Q, HU L, et al. Sequential/session-based recommendations: challenges, approaches, applications and opportunities[C]//Proceedings of the 45th International ACM SIGIR Conference on Research and Development in Information Retrieval, 2022: 3425-3428.
[2] 罗旭, 汪海涛, 贺建峰. 基于双通道轻量图卷积的序列推荐算法[J]. 计算机工程与科学, 2024, 46(3): 560-570.
LUO X, WANG H T, HE J F. Sequential recommendation based on dual-channel light graph convolution[J]. Computer Engineering & Science, 2024, 46(3): 560-570.
[3] 陈万志, 王军. 时间感知增强的动态图神经网络序列推荐算法[J]. 计算机工程与应用, 2024, 60(20): 142-152.
CHEN W Z, WANG J. Time-aware enhancement dynamic graph neural networks for sequential recommendation algorithm[J]. Computer Engineering and Applications, 2024, 60(20): 142-152.
[4] KANG W C, MCAULEY J. Self-attentive sequential recommendation[C]//Proceedings of the 2018 IEEE International Conference on Data Mining, 2018: 197-206.
[5] 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, 2019: 1441-1450.
[6] TAN Q, ZHANG J, YAO J, et al. Sparse-interest network for sequential recommendation[C]//Proceedings of the 14th ACM International Conference on Web Search and Data Mining, 2021: 598-606.
[7] LI M, ZHANG Z, ZHAO X, et al. AutoMLP: automated MLP for sequential recommendations[C]//Proceedings of the 2023 ACM Web Conference, 2023: 1190-1198.
[8] YAN Y, HE Y, LI L. Why time flies? The role of immersion in short video usage behavior[J]. Frontiers in Psychology, 2023, 14: 1127210.
[9] ZHAN R, PEI C, SU Q, et al. Deconfounding duration bias in watch-time prediction for video recommendation[C]//Proceedings of the 28th ACM SIGKDD Conference on Knowledge Discovery and Data Mining, 2022: 4472-4481.
[10] LIN X, CHEN X, SONG L, et al. Tree based progressive regression model for watch-time prediction in short-video recommendation[J]. arXiv:2306.03392, 2023.
[11] RENDLE S, FREUDENTHALER C, SCHMIDT-THIE-ME L. Factorizing personalized Markov chains for next-basket recommendation[C]//Proceedings of the 19th International Conference on World Wide Web, 2010: 811-820.
[12] TANG J, WANG K. Personalized top-n sequential recom-mendation via convolutional sequence embedding[C]//Proceedings of the 11th ACM International Conference on Web Search and Data Mining, 2018: 565-573.
[13] YUAN F, KARATZOGLOU A, ARAPAKIS I, et al. A simple convolutional generative network for next item recommendation[C]//Proceedings of the 12th ACM International Conference on Web Search and Data Mining, 2019: 582-590.
[14] HIDASI B, KARATZOGLOU A, BALTRUNAS L, et al. Session-based recommendations with recurrent neural networks[J]. arXiv:1511.06939, 2015.
[15] LIU Y, ZHANG Y, WANG Y, et al. A survey of visual transformers[J]. IEEE Transactions on Neural Networks and Learning Systems, 2023, 35(6): 7478-7498.
[16] ZHOU P, YE Q, XIE Y, et al. Attention calibration for transformer-based sequential recommendation[C]//Proceedings of the 32nd ACM International Conference on Information and Know-ledge Management, 2023: 3595-3605.
[17] ZHANG Y, WANG X, CHEN H, et al. Adaptive disentan-gled transformer for sequential recommendation[C]//Proceedings of the 29th ACM SIGKDD Conference on Know-ledge Discovery and Data Mining, 2023: 3434-3445.
[18] LI C, WANG Y, LIU Q, et al. STRec: sparse transformer for sequential recommendations[C]//Proceedings of the 17th ACM Conference on Recommender Systems, 2023: 101-111.
[19] TRAN V A, SALHA-GALVAN G, SGUERRA B, et al. Attention mixtures for time-aware sequential recommendation[C]//Proceedings of the 46th International ACM SIGIR Conference on Research and Development in Information Retrieval, 2023: 1821-1826.
[20] ZHANG Y, BAI Y, CHANG J, et al. Leveraging watch-time feedback for short-video recommendations: a causal labeling framework[C]//Proceedings of the 32nd ACM International Conference on Information and Knowledge Management, 2023: 4952-4959.
[21] CAI Q, LIU S, WANG X, et al. Reinforcing user retention in a billion scale short video recommender system[J]. arXiv:2302.01724, 2023.
[22] CAI Q, XUE Z, ZHANG C, et al. Two-stage constrained actor-critic for short video recommendation[C]//Proceedings of the 2023 ACM Web Conference, 2023: 865-875.
[23] LI D, 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, 2020: 1553-1556.
[24] ZHENG Y, GAO C, DING J, et al. DVR: micro-video recommendation optimizing watch-time-gain under duration bias[C]//Proceedings of the 30th ACM International Conference on Multimedia, 2022: 334-345.
[25] DRAPER D, GUO E. The practical scope of the central limit theorem[J]. arXiv:2111.12267, 2021.
[26] WU C, WU F, QI T, et al. FeedRec: news feed recommendation with various user feedbacks[C]//Proceedings of the 2022 ACM Web Conference, 2022: 2088-2097.
[27] VASWANI A, SHAZEER N, PARMAR N, et al. Attention is all you need[C]//Proceedings of the 31st International Conference on Neural Information Processing Systems, 2017: 5998-6008.
[28] GAO C, LI S, ZHANG Y, et al. KuaiRand: an unbiased sequential recommendation dataset with randomly exposed videos[C]//Proceedings of the 31st ACM International Conference on Information & Knowledge Management, 2022: 3953-3957.
[29] RENDLE S, FREUDENTHALER C, GANTNER Z, et al. BPR: Bayesian personalized ranking from implicit feedback[J]. arXiv:1205.2618, 2012.
[30] DEVLIN J, CHANG M W, LEE K, et al. BERT: pretraining of deep bidirectional transformers for language understanding[J]. arXiv:1810.04805, 2018.
[31] FAN X, LIU Z, LIAN J, et al. Lighter and better: low-rank decomposed self-attention networks for next-item recommendation[C]//Proceedings of the 44th International ACM SIGIR Conference on Research and Development in Information Retrieval, 2021: 1733-1737.
[32] DU X, YUAN H, ZHAO P, et al. Frequency enhanced hybrid attention network for sequential recommendation[C]//Proceedings of the 46th International ACM SIGIR Conference on Research and Development in Information Retrieval, 2023: 78-88.
[33] 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, 2021: 4653-4664.
[34] KINGMA D P, BA J. Adam: a method for stochastic opti-mization[J]. arXiv:1412.6980, 2014.