Friendship Prediction Method Integrating Multi-view Information

doi:10.3778/j.issn.1673-9418.2412016

Abstract

Abstract: Location-based social networks (LBSNs), which contain the friend relationships and the trajectories of users checking in POIs, have been widely used to recommend POIs to users, plan events for users and predict the friend relationships between users. However, existing methods for predicting the friend relationships usually ignore the users?? preferences, spatial constraints and dynamic changes for POIs hidden in the check-in trajectories, which have important effects on friend relationship prediction. To address the above issues, a friendship prediction method integrating multi-view information (FPIMV) is proposed to enhance the embeddings of user pairs and improve friendship prediction by capturing the spatial-temporal information in the check-in trajectories and the friend relationships between users. Specifically, in the spatial information embedding layer of FPIMV, this paper proposes a method for learning the embeddings of user pairs based on the characteristics of check-in POIs, which effectively captures the users?? preferences and spatial constraints for POIs by learning the similarity and the proximity of POIs. In the temporal information embedding layer of FPIMV, this paper proposes a method for learning the embeddings of user pairs based on the characteristics of check-in time, which effectively captures the users?? dynamic changes for POIs by learning the local and global features of check-in time. In the social relationship embedding layer of FPIMV, this paper learns the embeddings of user pairs based on the meta-paths, which capture the semantic relationships between users. Finally, this paper enhances the embeddings of user pairs by fusing the above embeddings learnt. Experimental results on three datasets from LBSNs show that compared with the baselines, FPIMV improves the performance of friendship prediction.

Key words: location-based social networks, friendship prediction, network embedding, spatial-temporal information

摘要： 位置社会网络（LBSN）是一类融合了朋友关系和用户签到轨迹的社会网络，其广泛应用于兴趣点推荐、活动策划和朋友关系预测等任务。然而，现有的朋友关系预测方法在处理用户签到轨迹时，在空间维度上忽略了用户的兴趣点偏好和空间约束，在时间维度上忽略了用户兴趣点偏好的动态变化。针对上述问题，提出融合多视角信息的朋友关系预测方法（FPIMV），旨在通过捕获用户签到轨迹中的空间、时间信息以及用户间的朋友关系等多视角信息，有效学习用户对嵌入，提升朋友关系预测效果。FPIMV模型在签到位置嵌入层，提出基于兴趣点相似性和邻近度的用户对嵌入方法，该方法有效捕获了用户的兴趣点偏好和空间约束；在签到时间嵌入层，提出基于签到时间局部和全局特征的用户对嵌入方法，该方法有效捕获了用户兴趣点偏好的动态变化；在社会关系嵌入层，基于元路径学习用户对嵌入，以捕获用户间的语义关系；最后，融合上述嵌入，提高用户对嵌入质量，进而提升朋友关系预测性能。在三个LBSN数据集上的实验结果表明，FPIMV模型在朋友关系预测任务上优于基线模型。

关键词: 位置社会网络, 朋友关系预测, 网络嵌入, 时空信息

MA Zhihui, CHEN Hongmei, YANG Peizhong, WANG Lizhen. Friendship Prediction Method Integrating Multi-view Information[J]. Journal of Frontiers of Computer Science and Technology, 2025, 19(11): 3023-3032.

马志慧, 陈红梅, 杨培忠, 王丽珍. 融合多视角信息的朋友关系预测方法[J]. 计算机科学与探索, 2025, 19(11): 3023-3032.

References

[1] CHANG S, TAO Y T, ZHU H Z, et al. FriendSeeker: inferring hidden friendship in mobile social networks with sparse check-in data[C]//Proceedings of the 2023 IEEE 43rd International Conference on Distributed Computing Systems. Piscataway: IEEE, 2023: 440-450.
[2] QIN Y F, WU H J, JU W, et al. A diffusion model for POI recom-mendation[J]. ACM Transactions on Information Systems, 2024, 42(2): 1-27.
[3] 李曼文, 张月琴, 张晨威, 等. 异质图嵌入的地理不敏感时空兴趣点推荐方法[J]. 计算机科学与探索, 2024, 18(3): 755-767.
LI M W, ZHANG Y Q, ZHANG C W, et al. Geographically insensitive spatial-temporal POI recommendation based on heterogeneous graph embedding[J]. Journal of Frontiers of Computer Science and Technology, 2024, 18(3): 755-767.
[4] WU Y J, LIAN D F, JIN S W, et al. Graph convolutional networks on user mobility heterogeneous graphs for social relationship inference[C]//Proceedings of the 28th International Joint Conference on Artificial Intelligence. Palo Alto: AAAI, 2019: 3898-3904.
[5] ZHANG W, LAI X, WANG J Y. Social link inference via multiview matching network from spatiotemporal trajectories[J]. IEEE Transactions on Neural Networks and Learning Systems, 2023, 34(4): 1720-1731.
[6] TRUNG H, VINH T, NGUYEN T, et al. MSC-lbsn: multi-social context-aware hypergraph representation learning for lbsns[J]. lEEE Transactions on Knowledge and Data Engineering, 2022, 2: 2696407048.
[7] LI Y, FAN Z, ZHANG J, et al. Heterogeneous hypergraph neural network for friend recommendation with human mobility[C]//Proceedings of the 31st ACM International Conference on Information & Knowledge Management. New York: ACM, 2022: 4209-4213.
[8] GROVER A, LESKOVEC J. node2vec: scalable feature learning for networks[C]//Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. New York: ACM, 2016: 855-864.
[9] VELI?KOVI? P, CUCURULL G, CASANOVA A, et al. Graph attention networks[EB/OL]. [2024-10-12]. https://arxiv.org/abs/1710.10903.
[10] BACKES M, HUMBERT M, PANG J, et al. walk2friends: inferring social links from mobility profiles[C]//Proceedings of the 2017 ACM SIGSAC Conference on Computer and Communications Security. New York: ACM, 2017: 1943-1957.
[11] YU Y W, WANG H J, LI Z H. Inferring mobility relationship via graph embedding[J]. Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies, 2018, 2(3): 1-21.
[12] ZHOU F, WU B Y, YANG Y, et al. Vec2Link: unifying heterogeneous data for social link prediction[C]//Proceedings of the 27th ACM International Conference on Information and Knowledge Management. New York: ACM, 2018: 1843-1846.
[13] FENG J, ZHANG M Y, WANG H D, et al. DPLink: user identity linkage via deep neural network from heterogeneous mobility data[C]//Proceedings of the 2019 World Wide Web Conference. New York: ACM, 2019: 459-469.
[14] YANG D, QU B, YANG J, et al. Revisiting user mobility and social relationships in LBSNs: a hypergraph embedding approach[C]//Proceedings of the 2019 World Wide Web Conference. New York: ACM, 2019: 2147-2157.
[15] YANG D Q, QU B Q, YANG J, et al. LBSN2Vec++: heterogeneous hypergraph embedding for location-based social networks[J]. IEEE Transactions on Knowledge and Data Engineering, 2020, 34(4): 1843-1855.
[16] LI Y K, FAN Z P, YIN D, et al. HMGCL: heterogeneous multigraph contrastive learning for LBSN friend recommendation[J]. World Wide Web, 2023, 26(4): 1625-1648.
[17] QIN G M, SONG L X, YU Y W, et al. Graph structure learning on user mobility data for social relationship inference[C]//Proceedings of the 37th AAAI Conference on Artificial Intelligence and the 35th Conference on Innovative Applications of Artificial Intelligence and the 13th Symposium on Educational Advances in Artificial Intelligence. Palo Alto: AAAI, 2023: 4578-4586.
[18] QIN G M, QI J P, WANG B, et al. Multi-relational graph attention network for social relationship inference from human mobility data[C]//Proceedings of the 33rd International Joint Conference on Artificial Intelligence. Palo Alto: AAAI, 2024: 2315-2323.
[19] DONG Y X, HU Z N, WANG K S, et al. Heterogeneous network representation learning[C]//Proceedings of the 29th International Joint Conference on Artificial Intelligence. Palo Alto: AAAI, 2020: 4861-4867.
[20] MARIA E, BUDIMAN E, HAVILUDDIN, et al. Measure distance locating nearest public facilities using haversine and Euclidean methods[J]. Journal of Physics: Conference Series, 2020, 1450(1): 012080.
[21] SADHWANI D, POWARI A, MEHTA N. New, simple and accurate approximation for the Gaussian Q function with applications[J]. IEEE Communications Letters, 2022, 26(3): 518-522.
[22] VASWANI A, SHAZEER N, PARMAR N, et al. Attention is all you need[C]//Advances in Neural Information Processing Systems 30, 2017: 5998-6008.
[23] LIU M H, ZENG A L, CHEN M H, et al. SCINet: time series modeling and forecasting with sample convolution and interaction[C]//Advances in Neural Information Processing Systems 35, 2022: 5816-5828.