融合人群移动轨迹和时空-类别的下一个兴趣点推荐

doi:10.3778/j.issn.1002-8331.2401-0424

摘要/Abstract

摘要： 下一个兴趣点推荐（next POI recommendation）作为基于位置社交网络的主要应用之一，为用户和服务提供商带来了显著的实用价值。现有的POI推荐模型主要依赖于目标用户的历史签到数据进行推荐，没有充分利用其他用户移动轨迹数据的潜在价值，也未有效提取和融合时空-类别信息的特征。为了解决上述问题，提出了一种融合人群移动轨迹和时空-类别的下一个兴趣点推荐模型（GGCN-STC）。依据用户的移动轨迹构建区域轨迹图，提出了门控图卷积神经网络对共同移动轨迹进行建模；将签到序列中的时空-类别信息进行多维度的特征融合；利用自注意力机制捕获用户偏好，为用户提供更准确的POI推荐。在两个真实数据集上进行实验比较与分析，结果表明该模型优于其他模型。

关键词: 兴趣点推荐, 门控图卷积神经网络, 自注意力机制, 时空网络

Abstract: Next point of interest (POI) recommendation, as one of the main applications of location-based social networks, brings significant practical value to both users and service providers. However, existing POI recommendation models mainly rely on the historical check-in data of target users, without fully leveraging the potential value of movement trajectories data from others, nor effectively extracting and fusing spatiotemporal-category features. To address these issues, this paper proposes GGCN-STC, a next POI recommendation model that integrates crowd movement trajectories and spatiotemporal-category features. Firstly, a region trajectory graph is constructed based on user movement trajectories, and the gated graph convolutional neural network is proposed to construct a model of shared graph. Secondly, spatiotemporal-category information from check-in sequences is fused into multi-dimensional features. Finally, it provides more accurate POI recommendations for users by capturing their preferences with self-attention mechanisms. Experiments are conducted for comparison and analysis on two real datasets, and the results demonstrate that the proposed model outperforms others.

Key words: POI recommendation, gated graph convolutional neural network, self-attention, spatiotemporal network

郭秉璇, 杨晓文, 孙福盛, 况立群, 张元, 韩慧妍. 融合人群移动轨迹和时空-类别的下一个兴趣点推荐[J]. 计算机工程与应用, 2025, 61(2): 316-326.

GUO Bingxuan, YANG Xiaowen, SUN Fusheng, KUANG Liqun, ZHANG Yuan, HAN Huiyan. Next Point of Interest Recommendation Fusing Crowd Movement Trajectories and Spatiotemporal-Category Features[J]. Computer Engineering and Applications, 2025, 61(2): 316-326.

参考文献

[1] ISLAM M A, MOHAMMAD M M, DAS S S S, et al. A survey on deep learning based point-of-interest recommendations[J]. Neurocomputing, 2022, 472: 306-325.
[2] YAN X D, SONG T W, JIAO Y F, et al. Spatio-temporal hypergraph learning for next poi recommendation[C]//Proceedings of the 46th International Conference on Research and Development in Information Retrieval, 2023: 403-412.
[3] LIU Q, WU S, WANG L, et al. Predicting the next location: a recurrent model with spatial and temporal contexts[C]//Proceedings of the 13th AAAI Conference on Artificial Intelligence, 2016: 194-200.
[4] LENG Z Q, LIU Y H, ZHOU X, et al. A long and?short term preference model for?next point of?interest recommendation[C]//Proceedings of the 31st International Conference on Artificial Neural Networks, 2022: 744-756.
[5] LIU Y W, PEI A X, WANG F, et al. An attention-based category-aware GRU model for the next poi recommendation[J].International Journal of Intelligent Systems, 2021, 36(7): 3174-3189.
[6] 柴瑞敏, 殷臣.用户关系和上下文感知的下一个兴趣点推荐[J].计算机工程与应用, 2022, 58(7): 197-205.
CHAI R M, YIN C. User relationship and context-aware next point of interest recommendation[J]. Computer Engineering and Applications, 2022, 58(7): 197-205.
[7] XIE J Y, CHEN Z Z. Hierarchical transformer with spatio-temporal context aggregation for next point-of-interest recommendation[J]. ACM Transactions on Information Systems, 2023, 42(2): 1-30.
[8] 刘树越, 于亚新, 吴晓露, 等.自注意力下时空-语义相融合的POI序列推荐[J].小型微型计算机系统, 2023, 44(3): 456-462.
LIU S Y, YU Y X, WU X L, et al. POI sequence recommendation model based on the integration of spatiotemporal and semantics under self-attention[J].Journal of Chinese Computer Systems, 2023, 44(3): 456-462.
[9] XU X H, SUZUMURA T, YONG J W, et al. Revisiting mobility modeling with graph: a graph transformer model for next point-of-interest recommendation[C]//Proceedings of the 31st ACM International Conference on Advances in Geographic Information Systems, 2023: 1-10.
[10] HUANG H S, GARTNER G. Using trajectories for collaborative filtering-based poi recommendation[J]. International Journal of Data Mining, Modelling and Management, 2014, 6(4): 333-346.
[11] GAO R, LI J, LI X F, et al. A personalized point-of-interest recommendation model via fusion of geo-social information[J]. Neurocomputing, 2018, 273: 159-170.
[12] LI H Y, GE Y, HONG R C, et al. Point-of-interest recommendations: learning potential check-ins from friends[C]//Proceedings of the 22nd International Conference on Knowledge Discovery and Data Mining, 2016: 975-984.
[13] 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, 2010: 811-820.
[14] WANG X F, FUKUMOTO F, CUI J, et al. EEDN: enhanced encoder-decoder network with local and global context learning for POI recommendation[C]//Proceedings of the 46th International Conference on Research and Development in Information Retrieval, 2023: 383-392.
[15] FANG J F, MENG X F. URPI-GRU: an approach of next poi recommendation based on user relationship and preference information[J]. Knowledge-Based Systems, 2022, 256: 109848.
[16] 李鹏飞, 贺洋, 毋建宏. 融合全局特征的时空网络兴趣点推荐算法[J].计算机工程与应用, 2024, 60(11): 75-83.
LI P F, HE Y, WU J H. A spatio-temporal network interest point recommendation algorithm fusing global features[J]. Computer Engineering and Applications, 2024, 60(11): 75-83.
[17] YIN F Y, LIU Y, SHEN Z Q, et al. Next POI recommendation with dynamic graph and explicit dependency[C]//Proceedings of the 37th AAAI Conference on Artificial Intelligence and 35th Conference on Innovative Applications of Artificial Intelligence and 13th Symposium on Educational Advances in Artificial Intelligence, 2023: 4827-4834.
[18] LI Y J, TARLOW D, BROCKSCHMIDT M, et al. Gated graph sequence neural networks[C]//Proceedings of the 4th International Conference on Learning Representations, 2016: 6303-6318.
[19] ZHAO P P, LUO A J, LIU Y C, et al. Where to go next: a spatio-temporal gated network for next poi recommendation[J]. IEEE Transactions on Knowledge and Data Engineering, 2022, 34(5): 2512-2524.
[20] LIN Y, WAN H Y, GUO S N, et al. Pre-training context and time aware location embeddings from spatial-temporal trajectories for user next location prediction[C]//Proceedings of the AAAI Conference on Artificial Intelligence, 2021: 4241-4248.
[21] KAZEMI S M, GOEL R, EGHBALI S, et al. Time2vec: learning a vector representation of time[J]. arXiv:1907.05321, 2019.
[22] 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: 6000-6010.
[23] YANG S, LIU J M, ZHAO K Q. GETNext: trajectory flow map enhanced transformer for next poi recommendation[C]//Proceedings of the 45th International Conference on Research and Development in Information Retrieval, 2022: 1144-1153.
[24] LUO Y T, LIU Q, LIU Z C. STAN: spatio-temporal attention network for next location recommendation[C]//Proceedings of the Web Conference, 2021: 2177-2185.
[25] WANG Z W, ZENG J, WEN J H, et al. Point-of-interest recommendation using deep semantic model[J]. Expert Systems with Applications, 2023, 231: 120727.