基于意图的轻量级自注意力序列推荐模型

doi:10.3969/j.issn.1006-2475.2024.12.001

摘要/Abstract

摘要： 现有序列推荐模型中自注意力机制计算参数量过大，同时没有充分考虑用户购物意图中的偏好信息。本文提出一种基于意图的轻量级自注意力序列推荐模型。该模型在传统的商品序列编码基础上，引入意图序列编码，进一步挖掘序列间转换模式；同时，为了降低序列中两两商品/意图间自注意力计算复杂度，设计卷积分段采样模块，将用户行为序列和意图序列分为多个片段，即将用户兴趣映射到多个序列片段中，应用自注意力机制捕捉片段间依赖关系，有效减少计算参数量。在MovieLens-1M、Yelp和Amazon-Books这3个公开数据集上进行对比实验，结果表明，相比基线模型，其命中率、归一化折损累计增益和平均倒数排名在MovieLens-1M数据集上提升了5.32%、4.40%和5.51%，在Yelp数据集上提升了30.93%、22.73%和28.84%，在Amazon-Books数据集上提升了7.78%、11.55%和13.98%，充分验证了本文所提模型的有效性。

关键词: 序列推荐, 意图推荐, 卷积神经网络, 自注意力机制

Abstract: The parameters of the self-attention calculation mechanism in the existing sequence recommendation models are too large， and there is insufficient preference information in the user's shopping intention. This paper proposes an intent-based lightweight self-attention network for sequential recommendation. On the basis of the traditional product sequence embedding， the model introduces intention sequence embedding to further explore the conversion patterns between sequences. At the same time， in order to reduce the computational complexity of self-attention between pairwise products/intentions in the sequence， a convolutional segmentation sampling module is designed to divide the user behavior sequence and intention sequence into multiple segments， mapping user interests to multiple sequence segments. Comparative experiments are conducted on three public datasets， MovieLens-1M， Yelp， and Amazon-Books. Compared with baseline models， the self-attention mechanism is applied to capture the dependency between segments， effectively reducing the number of parameters. The results show that the hit rate， normalized discounted cumulative gain and mean reciprocal ranking are increased by 5.32%， 4.40% and 5.51% on the MovieLens-1M dataset， 30.93%， 22.73% and 28.84% on the Yelp dataset， and 7.78%， 11.55% and 13.98% on the Amazon-Books dataset， which verify the effectiveness of the model proposed in this paper.

Key words: sequence recommendation； intent recommendation； convolution neural network； self-attention mechanisms ,

中图分类号:

TP391

何思达, 陈平华. 基于意图的轻量级自注意力序列推荐模型[J]. 计算机与现代化, 2024, 0(12): 1-9.

HE Sida, CHEN Pinghua. Intent-based Lightweight Self-Attention Network for Sequential Recommendation[J]. Computer and Modernization, 2024, 0(12): 1-9.

参考文献

［1］汪菁瑶，吴国栋，范维成，等. 用户行为序列个性化推荐研究综述［J］. 小型微型计算机系统， 2022，43（5）：921-935.
［2］ WANG S J， 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. ACM， 2022：3425-3428.
［3］ WANG S J， HU L， WANG Y， et al. Sequential recommender systems： Challenges， progress and prospects［J］. arXiv preprint arXiv：2001.04830， 2019.
［4］ CEN Y K， ZHANG J W， ZOU X， et al. Controllable multi-interest framework for recommendation［C］// Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining. ACM， 2020：2942-2951.
［5］ RENDLE S， FREUDENTHALER C， SCHMIDT-THIEME L. Factorizing personalized Markov chains for next-basket recommendation［C］// Proceedings of the 19th International Conference on World Wide Web. ACM， 2010：811-820.
［6］ WU C Y， AHMED A， BEUTEL A， et al. Recurrent recommender networks［C］// Proceedings of the 10th ACM International Conference on Web Search and Data Mining. ACM， 2017：495-503.
［7］ CHO K， VAN MERRIËNBOER B， GULCEHRE C， et al. Learning phrase representations using RNN encoder-decoder for statistical machine translation［J］. arXiv preprint arXiv：1406.1078， 2014.
［8］ HIDASI B， KARATZOGLOU A， BALTRUNAS L， et al. Session-based recommendations with recurrent neural networks［J］. arXiv preprint arXiv：1511.06939， 2015.
［9］ SONG Y， LEE J G. Augmenting recurrent neural networks with high-order user-contextual preference for session-based recommendation［J］. arXiv preprint arXiv：1805.02983，
2018.
［10］ KUMAR V， KHATTAR D， GUPTA S， et al. Deep neural architecture for news recommendation［C］// The 8th International Conference of the CLEF Initiative， CEUR Workshop Proceedings. 2017.
［11］ DAI H J， WANG Y C， TRIVEDI R， et al. Deep coevolutionary network： Embedding user and item features for recommendation［J］. arXiv preprint arXiv：1609.03675， 2016.
［12］ VASSØY B， RUOCCO M， DE SOUZA DA SILVA E， et al. Time is of the essence： A joint hierarchical RNN and point process model for time and item predictions［C］// Proceedings of the 12th ACM International Conference on Web Search and Data Mining. ACM，2019：591-599.
［13］ TANG J X， WANG K. Personalized top-n sequential recommendation via convolutional sequence embedding［C］// Proceedings of the 11th ACM International Conference on Web Search and Data Mining. ACM， 2018：565-573.
［14］ HSU K C， CHOU S Y， YANG Y H， et al. Neural network based next-song recommendation［J］. arXiv preprint arXiv：1606.07722， 2016.
［15］ TUAN T X， PHUONG T M. 3D convolutional networks for session-based recommendation with content features［C］//Proceedings of the 11th ACM Conference on Recommender Systems. ACM， 2017：138-146.
［16］ YAN A， CHENG S， KANG W C， et al. CosRec：2D convolutional neural networks for sequential recommendation［C］// Proceedings of the 28th ACM International Conference on Information and Knowledge Management. ACM， 2019：2173-2176.
［17］ YUAN F J， HE X N， JIANG H C， et al. Future data helps training： Modeling future contexts for session-based recommendation［C］// Proceedings of The Web Conference 2020. ACM， 2020：303-313.
［18］ YU F， KOLTUN V. Multi-scale context aggregation by dilated convolutions［J］. arXiv preprint arXiv：1511.07122， 2015.
［19］ YAKHCHI S， BEHEHSTI A， GHAFARI S M， et al. A convolutional attention network for unifying general and sequential recommenders［J］. Information Processing & Management， 2022， 59（1）. DOI: 10.1016/j.ipm.2020.102755.
［20］ 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. Curran Associates Inc， 2017：6000-6010.
［21］ KANG W C， MCAULEY J. Self-attentive sequential recommendation［C］//2018 IEEE International Conference on Data Mining （ICDM）. IEEE， 2018：197-206.
［22］ ZHANG S， TAY Y， YAO L N， et al. Next item recommendation with self-attention［J］. arXiv preprint arXiv：1808.
06414， 2018.
［23］ LI J C， WANG Y J， MCAULEY J. Time interval aware self-attention for sequential recommendation［C］// Proceedings of the 13th International Conference on Web Search and Data Mining. ACM， 2020：322-330.
［24］ ALAPARTHI S， MISHRA M. Bidirectional encoder representations from transformers （BERT）： A sentiment analysis odyssey［J］. arXiv preprint arXiv：2007.01127， 2020.
［25］ 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. ACM， 2019：1441-1450.
［26］ WU L W， LI S Q， HSIEH C J， et al. SSE-PT： Sequential recommendation via personalized transformer［C］// Proceedings of the 14th ACM Conference on Recommender Systems. ACM， 2020：328-337.
［27］ FAN X Y， LIU Z， LIAN J X， 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.
［28］ CHEN W Y， REN P J， CAI F， et al. Improving end-to-end sequential recommendations with intent-aware diversification［C］// Proceedings of the 29th ACM International Conference on Information & Knowledge Management. ACM， 2020：175-184.
［29］ TANJIM M， SU C Z， BENJAMIN E， et al. Attentive sequential models of latent intent for next item recommendation［C］// Proceedings of The Web Conference2020. ACM， 2020：2528-2534.
［30］ LI H Y， WANG X， ZHANG Z W， et al. Intention-aware sequential recommendation with structured intent transition［J］. IEEE Transactions on Knowledge and Data Engineering， 2022，34（11）：5403-5414.
［31］ ZHAO W X， HOU Y P， PAN X Y， 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. ACM， 2022： 4722-4726.

[1]	张晓东1, 白广芝1, 李敏1, 李昊洋2. 基于经验小波变换的油气井产量预测模型 [J]. 计算机与现代化, 2024, 0(12): 53-58.
[2]	刘宝宝, 杨菁菁, 陶露, 王贺应. 基于注意力的DSMSC的遥感图像场景分类[J]. 计算机与现代化, 2024, 0(12): 72-77.
[3]	祁贤, 刘大铭, 常佳鑫. 基于改进自注意力机制的多视图三维重建[J]. 计算机与现代化, 2024, 0(11): 106-112.
[4]	陈雪松1, 李衡1, 王浩畅2. 结合注意力机制和Mengzi模型的短文本分类[J]. 计算机与现代化, 2024, 0(09): 101-106.
[5]	高帅鹏, 王怡凡. 基于图像的群体情绪识别综述[J]. 计算机与现代化, 2024, 0(08): 98-107.
[6]	周宪溪, 牟莉. 基于改进TF-IDF和AGLCNN的新闻长文本分类模型[J]. 计算机与现代化, 2024, 0(08): 120-126.
[7]	杨江1, 孙晓梅1, 许韬2. 基于业务内容构建股票关联关系的股价预测[J]. 计算机与现代化, 2024, 0(07): 21-25.
[8]	刘存莉1, 雷占占2, 郑澳2. 基于循环卷积神经网络的排水管网缺陷检测方法[J]. 计算机与现代化, 2024, 0(07): 26-35.
[9]	李珊, 王林娜, 高丁佳, 宣海波. 基于图神经网络的多层银企网络融合研究[J]. 计算机与现代化, 2024, 0(05): 27-32.
[10]	钟海龙1, 2, 何月顺1, 何璘琳1, 陈杰1, 田鸣3, 郑瑞银4. 基于代价敏感卷积神经网络的加密流量分类#br# #br#[J]. 计算机与现代化, 2024, 0(05): 55-60.
[11]	高埂1, 肖风丽2, 杨飞1. 基于改进MobileNetV3-Small的色素减退性皮肤病诊断[J]. 计算机与现代化, 2024, 0(05): 120-126.
[12]	游嘉靖1, 2, 何月顺1, 何璘琳1, 钟海龙1, 2. 基于AHP-CNN的加密流量分类方法[J]. 计算机与现代化, 2024, 0(04): 83-87.
[13]	许跃雯1, 李明1, 李莉2. 基于对比学习MocoV2的COVID-19图像分类#br#[J]. 计算机与现代化, 2024, 0(02): 81-87.
[14]	周成诚, 曾庆军, 杨康, 胡家铭, 韩春伟. 基于高效通道注意力模块的运动想象脑电识别[J]. 计算机与现代化, 2023, 0(12): 19-23.
[15]	刘付琪, 张达, 宋建华, 王海东. 基于CNN-BiLSTM的液压系统故障诊断[J]. 计算机与现代化, 2023, 0(09): 10-19.