一种基于增强结构嵌入的时序链路预测算法

An enhanced structural embedding algorithm for temporal link prediction

摘要

本文提出一种基于增强结构嵌入的时序链接预测算法(enhanced structural embedding algorithm for temporal link prediction, SE-TLP). SE-TLP设计双通道结构特征提取机制用于建模节点和连边特性,其联合4种加权节点中心性衡量节点地位,采用连边关系强度表示拓扑结构,并融合节点和连边特征以增强结构嵌入表示,提高预测准确性.同时,结合图卷积网络和门控循环单元,沿时间维度学习网络结构的演化.实验表明,SE-TLP在两类评价指标上均优于主流算法,性能分别提升7.47%和4.43%.

Abstract

In this paper, an enhanced structural embedding algorithm for temporal link prediction (SE-TLP) is proposed. SE-TLP employs a dual-channel structural feature extraction mechanism to separately model node and edge features. It integrates four weighted node centralities to assess node status, designs edge relation strength to represent topological structure, and fuses node and edge features to enhance structural embedding representations, thereby improving prediction accuracy. Meanwhile, SE-TLP combines a graph convolutional network with a gated recurrent unit to learn the network's structural evolution over time. Experiments show that SE-TLP outperforms mainstream algorithms on both evaluation metrics, achieving performance improvements of 7.47% and 4.43% respectively.

关键词

时序链路预测 / 网络嵌入 / 图卷积网络 / 门控循环单元(GRU) / 动态网络

Key words

temporal link prediction / network embedding / graph convolutional network / gated recurrent unit (GRU) / dynamic network

参考文献

[1] Artime O, Grassia M, De Domenico M, et al. Robustness and resilience of complex networks[J]. Nature Reviews Physics, 2024, 6(2): 114-131.

[2] Bloch F, Jackson M O, Tebaldi P . Centrality measures in networks[J]. Social Choice and Welfare, 2023, 61(2): 413-453.

[3] Zhang T, Zhang K, Lv L S, et al. Temporal link prediction using node centrality and time series[J]. International Journal of Future Computer and Communication, 2020, 9(3): 62-65.

[4] Choudhury N, Uddin S . Evolutionary features for dynamic link prediction in social networks[J]. Applied Sciences, 2023, 13(5): 2913.

[5] Maritan C A, Lee G K . Resource allocation and strategy[J]. Journal of Management, 2017, 43(8): 2411-2420.

[6] Xue G T, Zhong M, Li J X, et al. Dynamic network embedding survey[J]. Neurocomputing, 2022, 472: 212-223.

[7] Defferrard M, Bresson X, Vandergheynst P . Convolutional neural networks on graphs with fast localized spectral filtering[C]// Proceedings of the 30th International Conference on Neural Information Processing Systems. [S. l.]: ACM, 2016: 3844-3852.

[8] Mei P, Zhao Y H . Dynamic network link prediction with node representation learning from graph convolutional networks[J]. Scientific Reports, 2024, 14(1): 538.

[9] Nosouhian S, Nosouhian F, Khoshouei E A K . A review of recurrent neural network architecture for sequence learning: Comparison between LSTM and GRU[EB/OL]. [2025-05-13]. https://www.preprints.org/manuscript/202107.0252.

[10] Pareja A, Domeniconi G, Chen J, et al. EvolveGCN: evolving graph convolutional networks for dynamic graphs[J]. Proceedings of the AAAI Conference on Artificial Intelligence, 2020, 34(4): 5363-5370.

[11] Hardin J S, Sarkis G, Urc P C . Network analysis with the enron email corpus[J]. Journal of Statistics Education, 2015, 23(2): 1-27.

[12] Panzarasa P, Opsahl T, Carley K M . Patterns and dynamics of users' behavior and interaction: Network analysis of an online community[J]. Journal of the American Society for Information Science and Technology, 2009, 60(5): 911-932.

[13] Leskovec J, Kleinberg J, Faloutsos C . Graphs over time: densification laws, shrinking diameters and possible explanations[C]// Proceedings of the Eleventh ACM SIGKDD International Conference on Knowledge Discovery in Data Mining. [S. l.]: ACM, 2005: 177-187.

[14] Yin H, Benson A R, Leskovec J, et al. Local higher-order graph clustering[C]// Proceedings of the 23rd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. [S. l.]: ACM, 2017: 555-564.

[15] Kumar S, Spezzano F, Subrahmanian V S, et al. Edge weight prediction in weighted signed networks[C]// 2016 IEEE 16th International Conference on Data Mining (ICDM). [S. l.]: IEEE, 2016: 221-230.

[16] Rozemberczki B, Allen C, Sarkar R . Multi-Scale attributed node embedding[J]. Journal of Complex Networks, 2021, 9(2): cnab014.

[17] Papadopoulos F, Aldecoa R, Krioukov D . Network geometry inference using common neighbors[J]. Physical Review E, 2015, 92(2): 022807.

[18] Lü L Y, Jin C H, Zhou T . Similarity index based on local paths for link prediction of complex networks[J]. Physical Review E, 2009, 80(4): 046122.

[19] Jung J W, Heo H S, Yu H J, et al. Graph attention networks for speaker verification[C]// ICASSP 2021 - 2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). [S. l.]: IEEE, 2021: 6149-6153.

[20] Chen J Y, Wang X K, Xu X H . GC-LSTM: graph convolution embedded LSTM for dynamic network link prediction[J]. Applied Intelligence, 2022, 52(7): 7513-7528.

[21] Sserwadda A, Ozcan A, Yaslan Y . Structural and topological guided GCN for link prediction in temporal networks[J]. Journal of Ambient Intelligence and Humanized Computing, 2023, 14(7): 9667-9675.

基金资助

国家自然科学基金(62072160)

河南省科技攻关项目(252102210141)

河南省国际科技合作项目(262102521055)