-
针对复杂网络中关键节点的识别、评估及排序问题，受物理系统中不同节点间信息的多维度、多层次相互影响过程的启发，提出了一种基于图卷积神经网络的多维参数的节点重要性评估方法．该方法结合了卷积神经网络自动学习的特性，综合考虑节点的内在特性、与邻近节点的交互关系以及其在整个网络中的功能角色，构建了一种新颖的关键节点识别框架，即多维参数控制图卷积网络（Multi-Parameter Control Graph Convolutional Networks, MPC-GCN）．通过卷积神经网络对节点及其邻居特征的逐层聚合，自动提取并综合节点的局部特性、全局特性及位置特性，实现对节点重要性的多维度评估，同时引入灵活的参数调整机制，允许调整不同维度信息对评估结果的影响权重，以适应不同网络结构的需求．为验证该方法的有效性，在随机生成的小型网络上验证了参数对模型的作用；并在8个大型网络上利用SIR模型进行仿真实验，并以M(R)值、Kendall相关系数、被传染节点占比及最大连通子图相对大小作为评价标准．结果表明，MPC-GCN方法在单调性、准确性、适用性及鲁棒性上都优于其他相关方法，能够显著区分不同节点的重要程度．该方法有效地克服了现有方法评估角度单一和适应性方面的局限性，提高了评估的全面性和适用性．This paper addresses the problem of identifying, evaluating, and ranking key nodes in complex networks by introducing a novel Multi-Parameter Control Graph Convolutional Network (MPC-GCN) for assessing node importance. Drawing inspiration from the multidimensional and hierarchical interactions between nodes in physical systems, this approach integrates the automatic feature learning capabilities of Graph Convolutional Networks (GCNs) with a comprehensive analysis of nodes’ intrinsic properties, their interactions with neighbors, and their roles within the broader network. The MPC-GCN model offers an innovative framework for key node identification, leveraging GCNs to iteratively aggregate node and neighbor features across layers. This process captures and combines local, global, and positional characteristics, enabling a more nuanced, multidimensional assessment of node importance. Moreover, the model incorporates a flexible parameter adjustment mechanism, allowing the relative weights of different dimensions to be tuned, thereby adapting the evaluation process to various network structures. To validate the model’s effectiveness, we first tested the influence of model parameters on randomly generated small networks. We then conducted extensive simulations on eight large-scale networks using the Susceptible-Infected-Recovered (SIR) model. Evaluation metrics, including the M(R) score, Kendall’s tau correlation, the proportion of infected nodes, and the relative size of the largest connected component, were used to assess the model’s performance. The results demonstrate that MPC-GCN outperforms existing methods in terms of monotonicity, accuracy, applicability, and robustness, providing more precise differentiation of node importance. By addressing the limitations of current methods—such as their reliance on single-dimensional perspectives and lack of adaptability—MPC-GCN offers a more comprehensive and flexible approach to node importance evaluation. This method significantly improves the breadth and applicability of node ranking in complex networks.
-
[1] Watts D J, Strogatz S H 1998 Nature 393 440
[2] Barabási A-L, Albert R 1999 Science 286 509
[3] Xu Y L, Guo T Y, Tang K, Zhang Y Y, Li L W 2024 Acta Armamentarii 45 552（in Chinese） [许怡岚，郭唐仪，唐坤，张滢颖，李林蔚 2024 兵工学报 45 552]
[4] Sun L N, Liang B H, Chen Z W 2022 Fire Control & Command Control 47 119（in Chinese） [孙利娜，梁葆华，陈志伟 2022 火力与指挥控制 47 119]
[5] Li X L, Han Y L, Wu X G, Zhang D Y 2018 Journal of YanShan University 42 444（in Chinese） [李晓龙，韩益亮，吴旭光，张德阳 2018 燕山大学学报 42 444]
[6] Luo H, Yan G H, Zhang M, Bao J B, Li J C, Liu T, Yang B, Wei J 2020 Journal of Computer Research and Development 57 954（in Chinese） [罗浩，闫光辉，张萌，包峻波，李俊成，刘婷，杨波，魏军 2020 计算机研究与发展 57 954]
[7] Klemm K, Serrano M Á, Eguíluz V M, Miguel M S 2012 Scientific Reports 2
[8] Wang L L, Huang M, Gao L 2020 Journal of Transport Information and Safety 38 80（in Chinese） [王灵丽，黄敏，高亮 2020 交通信息安全 38 80]
[9] Lai Q, Zhang H H 2022 Chinese Physics B 31 068905
[10] Howell N 1985 Canadian Journal of Sociology 10 209
[11] Freeman L C 1977 Sociometry 40 35
[12] Sabidussi G 1966 Psychometrika 31 581
[13] Zareie A, Sheikhahmadi A, Khamforoosh K 2018 Expert Systems with Applications 108 96
[14] Li H, Shang Q, Deng Y 2021 Chaos, Solitons & Fractals 143 110456
[15] Zareie A, Sheikhahmadi A 2018 Expert Systems with Applications 93 200
[16] Yu H, Liu Z, Li Y J, Ieee 2013 5th International Conference on Measuring Technology and Mechatronics Automation (ICMTMA) Hong Kong, PEOPLES R CHINA, Jan 16-17,2013 p1292-1295
[17] Fan Y N, Liu S Y, Bai Y G 2020 Mathematics in Practice and Theory 50 159 （in Chinese） [樊燕妮，刘三阳，白艺光 2020 数学的实践与认识 50 159]
[18] Ma L L, Ma C, Zhang H F, Wang B H 2016 Physica a-Statistical Mechanics and Its Applications 451 205
[19] Jiang Y, Yang S Q, Yan Y W, Tong T C, Dai J Y 2022 Chinese Physics B 31 058903
[20] Yang X, Xiao FY 2021 Knowledge-Based Systems 227 107198
[21] Shang Q, Deng Y, Cheng K H 2021 Information Sciences 577 162
[22] Ai D, Liu X L, Kang W Z, Li L N, Lv S Q, Liu Y 2022 Chinese Physics B 32 118902
[23] Ullah A, Wang B, Sheng J, Long J, Khan N, Sun Z 2021 Expert Systems with Applications 186 115778
[24] Zhang X L, Tang J X 2021 Computer Engineering 47 139 （in Chinese） [张宪立，唐建新 2021 计算机工程 47 139]
[25] Ruan Y R, Lao S Y, Tang J, Bai L, Guo Y M 2022 Acta Phys. Sin. 71 176401（in Chinese） [阮逸润，老松杨，汤俊，白亮，郭延明 2022 物理学报 71 176401]
[26] Xu K, Hu W, Leskovec Proc 7th International Conference on Learning Representations (ICLR) LA, USA, May 6-9, 2018 p1467-5463
[27] Cao L, Ding C F, Ma L R, Yan Z Y, You H, Hong A Q 2024 Journal of Frontiers of Computer Science and Technology （in Chinese）[曹璐，丁苍峰，马乐荣，延照耀，游浩，洪安琪 2024 计算机科学与探索]
[28] Kipf, T N, Welling, M 2017 5th International Conference on Learning Representations Toulon, France, April 24-26, 2017
[29] Maurya S K, Liu X, Murata T 2019 Proceedings of the 28th ACM International Conference on Information and Knowledge Management Beijing, China, November 3-7, 2019 p2149-2152
[30] Kumar M S, Liu X, Murata T 2021 ACM Trans. 15 1
[31] Goel D, Shen H, Tian H, Guo M Y 2024 Expert Systems with Applications 249 123636
[32] Qu H, Song Y, Li R, Li M 2023 Physica A 632 129339
[33] Ramachandran K, Rj T 2022 The 2022 Intelligent Computing for Sustainable Energy and Environment Hobart, Australia, February 18-20, 2022 p1-10
[34] Sun C, Li C, Lim X, Zheng T, Meng F, Rui X, Wanf Z 2023 Artificial Intelligence Review. 56 2263
[35] Xiong C, Li W, Liu Y, Wang M 2021 IEEE Signal Processing Letters 28 573
[36] Li Z, Xing Y Y, Huang J M, Wang H B, Gao J L, Yu G X 2021 Future Generation Computer Systems 116 145
[37] Zhao G, Jia P, Zhou A, Zhang B 2020 Neurocomputing 414 18
[38] Liu C, Cao T, Zhou L 2022 Knowledge-Based Systems 251 109220
[39] Chen W, Feng F, Wang Q, He X, Song C, Ling G, Zhang Y 2023 IEEE Transactions on Knowledge and Data Engineering 35 3500
[40] Li W, Li T, Nikougoftar E 2024 Chaos, Solitons & Fractals 187 0960
[41] Yu E Y, Wang Y P, Fu Y, Chen D B, Xie M 2020 Knowledge-Based Systems 198 105893
[42] Zhang L, Song H, Aletras N, Lu H 2022 Pattern Recognition 128 108661
[43] Han B, Wei Y, Kang L, Wang Q, Yang Y 2022 Frontiers in Physics 9 2296
[44] Yang X, Xiao FY 2021 Knowledge-Based Systems 227 107198
[45] Ullah A, Wang B, Sheng J, Long J, Khan N, Sun Z 2021 Expert Systems with Applications 186 115778
[46] Zhu S, Zhan J, Li X 2023 Scientific Reports 13 16404
[47] Shang Q, Deng Y, Cheng K H 2021 Information Sciences 577 162
[48] Yang S Q, Jiang Y, Tong T C, Yan Y W, Gan G S 2021 Acta Phys. Sin. 70 216401 （in Chinese） [杨松青，蒋沅，童天驰，严玉为，淦各升 2021 物理学报 70 216401]
计量
- 文章访问数: 23
- PDF下载量: 0
- 被引次数: 0
下载:
