摘 要：针对基于时间间隔的序列推荐模型存在的非线性特征提取不充分问题，提出了增强非线性特征提取的时间间隔感知序列推荐模型，改进了已有的推荐模型。用多层线性层代替传统的基于时间间隔的序列推荐模型中的前馈神经网络，增强模型对于深层次项目交互信息的捕捉能力。在三个公开数据集上验证了所提出模型的有效性。评估指标平均提高1.9%，最高提升5.2%。
关键词：深度学习；推荐算法；序列推荐；时间序列；多层感知机
DOI:10.19850/j.cnki.2096-4706.2022.07.021
中图分类号：TP391 文献标识码：A 文章编号：2096-4706（2022）07-0085-04
Time Interval Aware Sequence Recommendation of Enhancing Nonlinear Feature Extraction
NING Yulin
(School of Computer Science and Engineering, Anhui University of Science & Technology, Huainan 232001, China)
Abstract: Aiming at the problem of existing insufficient nonlinear feature extraction based on the time interval sequence recommendation model, this paper proposes a time interval aware sequence recommendation model of enhancing nonlinear feature extraction, and it improves the existing recommendation model. The feed forward neural network based on the traditional time interval sequence recommendation model is replaced by multi-layer linear layer to enhance the model’s ability to capture the deep level item interaction information. The validity of the proposed model is verified on three public datasets, and the evaluation metrics increased by 1.9% on average and 5.2% at the highest.
Keywords: deep learning; recommendation algorithm; sequence recommendation; time sequence; multilayer perceptron
参考文献：
[1] ZHANG S，YAO L N，SUN A X，et al. Deep Learning Based Recommender System: A Survey and New Perspectives [J].ACM Computing Surveys，2019，52（1）：1-38.
[2] FANG H，ZHANG D N，SHU Y H，et al. Deep Learning for Sequential Recommendation：Algorithms，Influential Factors， and Evaluations [J/OL].arXiv：1905.01997 [cs.IR].[2022-02-07].https:// arxiv.org/abs/1905.01997.
[3] WANG S J，CAO L B，WANG Y，et al. A Survey on Session-based Recommender Systems [J].ACM Computing Surveys， 2022，54（7）：1-38.
[4] LI J C，WANG Y J，MCAULEY J L. Time Interval Aware Self-Attention for Sequential Recommendation [EB/OL]. 2020：322- 330.[2022-02-03].https://www.xueshufan.com/publication/2996931760.
[5] CHEN Q W，ZHAO H，LI W，et al. Behavior SequenceTransformer for E-commerce Recommendation in Alibaba [C]// Proceedings of the 1st International Workshop on Deep Learning
Practice for High-Dimensional Sparse Data.New York：Association for Computing Machinery，2019：1-4.
[6] HARPER，F M，KONSTAN J A. The MovieLens Datasets： History and Context [J].ACMTransactions on Interactive Intelligent Systems，2015，5（4）：1-19.
作者简介：宁昱霖（1997 －），男，汉族，安徽宿州人，硕士研究生在读，研究方向：推荐系统。