Abstract
Graphs arise naturally in numerous situations, including social graphs, transportation graphs, web graphs, protein graphs, etc. One of the important problems in these settings is to identify which nodes are important in the graph and how they affect the graph structure as a whole. Betweenness centrality and closeness centrality are two commonly used node ranking measures to find out influential nodes in the graphs in terms of information spread and connectivity. Both of these are considered as shortest path based measures as the calculations require the assumption that the information flows between the nodes via the shortest paths. However, exact calculations of these centrality measures are computationally expensive and prohibitive, especially for large graphs. Although researchers have proposed approximation methods, they are either less efficient or suboptimal or both. We propose the first graph neural network (GNN) based model to approximate betweenness and closeness centrality. In GNN, each node aggregates features of the nodes in multihop neighborhood. We use this feature aggregation scheme to model paths and learn how many nodes are reachable to a specific node. We demonstrate that our approach significantly outperforms current techniques while taking less amount of time through extensive experiments on a series of synthetic and real-world datasets. A benefit of our approach is that the model is inductive, which means it can be trained on one set of graphs and evaluated on another set of graphs with varying structures. Thus, the model is useful for both static graphs and dynamic graphs.
Source code is available at https://github.com/sunilkmaurya/GNN_Ranking
- Giorgio Ausiello and Luigi Laura. 2016. Directed hypergraphs: Introduction and fundamental algorithm: A survey. Theoretical Computer Science 658, Part B (2016), 293--306. DOI:https://doi.org/10.1016/j.tcs.2016.03.016Google Scholar
- Alex Bavelas. 1950. Communication patterns in task-oriented groups. Journal of the Acoustical Society of America 22, 6 (1950), 725--730. DOI:https://doi.org/10.1121/1.1906679Google ScholarCross Ref
- Alex Bavelas. 1948. A mathematical model for group structures. Applied Anthropology 7, 3 (1948), 16--30. DOI:https://doi.org/10.17730/humo.7.3.f4033344851gl053Google Scholar
- Murray A. Beauchamp. 1965. An improved index of centrality. Behavioral Science 10, 2 (1965), 161--163. DOI:https://doi.org/10.1002/bs.3830100205Google ScholarCross Ref
- Rianne van den Berg, Thomas Kipf, and Max Welling. 2017. Graph convolutional matrix completion. arXiv:1706.02263. Retrieved from https://arxiv.org/abs/1706.02263Google Scholar
- E. Bergamini, H. Meyerhenke, and C. Staudt. 2015. Approximating betweenness centrality in large evolving networks. In Proceedings of the 17th Workshop on Algorithm Engineering and Experiments. Society for Industrial and Applied Mathematics, 133--146. DOI:https://doi.org/10.1137/1.9781611973754.12Google Scholar
- Michele Borassi and Emanuele Natale. 2019. KADABRA is an Adaptive Algorithm for Betweenness via Random Approximation. ACM Journal of Experimental Algorithmics 24, 1 (2019), 1.2:1–1.2:35. DOI:https://doi.org/10.1145/3284359Google ScholarDigital Library
- Stephen P. Borgatti. 1995. Centrality and AIDS. Connections 18, 1 (1995), 112--115.Google Scholar
- Stephen P. Borgatti. 2005. Centrality and network flow. Social Network 27, 1 (2005), 55--71. DOI:https://doi.org/10.1016/j.socnet.2004.11.008Google ScholarCross Ref
- Ulrik Brandes. 2001. A faster algorithm for betweenness centrality. The Journal of Mathematical Sociology 25, 2 (2001), 163--177. DOI:https://doi.org/10.1080/0022250x.2001.9990249Google ScholarCross Ref
- Ulrik Brandes and Christian Pich. 2007. Centrality estimation in large networks. International Journal of Bifurcation and Chaos in Applied Sciences and Engineering 17, 7 (2007), 2303--2318. DOI:https://doi.org/10.1142/S0218127407018403Google ScholarCross Ref
- S. Brin and L. Page. 1998. The anatomy of a large-scale hypertextual web search engine. Computer Networks and ISDN Systems 30, 1--7 (1998), 107--117. Retrieved from http://ilpubs.stanford.edu:8090/361/Google ScholarDigital Library
- Dirk Brockmann and Dirk Helbing. 2013. The hidden geometry of complex, network-driven contagion phenomena. Science 342, 6164 (2013), 1337--1342. DOI:https://doi.org/10.1126/science.1245200Google Scholar
- Christopher J. C. Burges. 2010. From RankNet to LambdaRank to LambdaMART: An Overview. Microsoft Research Technical Report MSR-TR-2010-82. Microsoft, Albuquerque, NM.Google Scholar
- Nicola De Cao and Thomas Kipf. 2018. MolGAN: An implicit generative model for small molecular graphs. In ICML Workshop, Theoretical Foundations and Applications of Deep Generative Models. Retrieved from https://arxiv.org/abs/1805.11973Google Scholar
- Deli Chen, Yankai Lin, Wei Li, Peng Li, Jie Zhou, and Xu Sun. 2019. Measuring and relieving the over-smoothing problem for graph neural networks from the topological view. Proceedings of AAAI conference on Artificial Intelligence 34 (2019), 3438--3445. Retrieved from http://arxiv.org/abs/1909.03211Google Scholar
- Jie Chen, Tengfei Ma, and Cao Xiao. 2018. FastGCN: Fast learning with graph convolutional networks via importance sampling. In Proceedings of the International Conference on Learning Representations. Retrieved from https://openreview.net/forum?id=rytstxWAW¬eId=ByU9EpGSf.Google Scholar
- Wei Chen, Tie-yan Liu, Yanyan Lan, Zhi-ming Ma, and Hang Li. 2009. Ranking measures and loss functions in learning to rank. In Proceedings of theAdvances in Neural Information Processing Systems 22, Y. Bengio, D. Schuurmans, J. D. Lafferty, C. K. I. Williams, and A. Culotta (Eds.). Curran Associates, Inc., 315--323. Retrieved from http://papers.nips.cc/paper/3708-ranking-measures-and-loss-functions-in-learning-to-rank.pdf.Google Scholar
- Edith Cohen, Daniel Delling, Thomas Pajor, and Renato F. Werneck. 2014. Computing classic closeness centrality, at scale. In Proceedings of the 2nd ACM Conference on Online Social Networks . ACM, 37--50. DOI:https://doi.org/10.1145/2660460.2660465Google Scholar
- Gabriele Corso, Luca Cavalleri, Dominique Beaini, Pietro Liò, and Petar Veličković. 2020. Principal neighbourhood aggregation for graph nets. In Advances in Neural Processing Systems 33 (NeurIPS'20).Google Scholar
- Pierluigi Crescenzi, Pierre Fraigniaud, and Ami Paz. 2020. Simple and fast distributed computation of betweenness centrality. In IEEE Conference on Computer Communications. DOI:https://doi.org/10.1109/INFOCOM41043.2020.9155354Google ScholarDigital Library
- Michaël Defferrard, Xavier Bresson, and Pierre Vandergheynst. 2016. Convolutional neural networks on graphs with fast localized spectral filtering. arxiv:1606.09375 Retrieved from http://arxiv.org/abs/1606.09375Google Scholar
- Ying Ding, Erjia Yan, Arthur Frazho, and James Caverlee. 2010. PageRank for ranking authors in co-citation networks. Journal of the American Society for Information Science and Technology 60, 11 (2010), 2229--2243. DOI:https://doi.org/10.1002/asi.v60:11Google ScholarCross Ref
- David K. Duvenaud, Dougal Maclaurin, Jorge Iparraguirre, Rafael Bombarell, Timothy Hirzel, Alan Aspuru-Guzik, and Ryan P. Adams. 2015. Convolutional networks on graphs for learning molecular fingerprints. In Proceedings of the 28th International Conference on Neural Information Processing Systems - Volume 2. 2224--2232. Reterieved from http://papers.nips.cc/paper/5954-convolutional-networks-on-graphs-for-learning-molecular-fingerprints.pdf.Google Scholar
- Nick Edmonds, Torsten Hoefler, and Andrew Lumsdaine. 2010. A space-efficient parallel algorithm for computing betweenness centrality in distributed memory. In Proceedings of the 2010 International Conference on High Performance Computing. 1--10. DOI:https://doi.org/10.1109/HIPC.2010.5713180.Google ScholarCross Ref
- David Eppstein and Joseph Wang. 2001. Fast approximation of centrality. In Proceedings of the 12th Annual ACM-SIAM Symposium on Discrete Algorithms. Society for Industrial and Applied Mathematics, 228--229. Retrieved from http://dl.acm.org/citation.cfm?id=365411.365449.Google Scholar
- Federico Errica, Marco Podda, Davide Bacciu, and Alessio Micheli. 2019. A fair comparison of graph neural networks for graph classification. arXiv:1912.09893.Google Scholar
- Changjun Fan, Li Zeng, Yuhui Ding, Muhao Chen, Yizhou Sun, and Zhong Liu. 2019. Learning to identify high betweenness centrality nodes from scratch: A novel graph neural network approach. In Proceedings of the 28th ACM International Conference on Information and Knowledge Management. ACM, 559--568. DOI:https://doi.org/10.1145/3357384.3357979Google ScholarDigital Library
- Massimo Franceschet. 2011. PageRank: Standing on the shoulders of giants. Communications of the ACM 54, 6 (2011), 92--101. DOI:https://doi.org/10.1145/1953122.1953146Google ScholarDigital Library
- Linton C. Freeman. 1978. Centrality in social networks conceptual clarification. Social Networks 1, 3 (1978), 215--239. DOI:https://doi.org/10.1016/0378-8733(78)90021Google ScholarCross Ref
- Linton C. Freeman. 1977. A set of measures of centrality based on betweenness. Sociometry 40, 1 (1977), 35--41. DOI:https://doi.org/10.2307/3033543Google ScholarCross Ref
- Robert Geisberger, Peter Sanders, and Dominik Schultes. 2008. Better approximation of betweenness centrality. In Proceedings of the Meeting on Algorithm Engineering & Expermiments. 90--100. Retrieved from http://dl.acm.org/citation.cfm?id=2791204.2791213.Google ScholarDigital Library
- Justin Gilmer, Samuel S. Schoenholz, Patrick F. Riley, Oriol Vinyals, and George E. Dahl. 2017. Neural message passing for quantum chemistry. In Proceedings of the 34th International Conference on Machine Learning - Volume 70. 1263--1272. Retrieved from http://dl.acm.org/citation.cfm?id=3305381.3305512.Google Scholar
- David F. Gleich. 2015. PageRank beyond the web. Society for Industrial and Applied Mathematics 57, 3 (2015), 321--363. DOI:https://doi.org/10.1137/140976649Google ScholarDigital Library
- O. Green, R. McColl, and D. A. Bader. 2013. A fast algorithm for streaming betweenness centrality. In Proceedings of the 2012 International Conference on Privacy, Security, Risk and Trust and 2012 International Confernece on Social Computing. 11--20. DOI:https://doi.org/10.1109/SocialCom-PASSAT.2012.37Google ScholarDigital Library
- Aric A. Hagberg, Daniel A. Schult, and Pieter J. Swart. 2008. Exploring network structure, dynamics, and function using NetworkX. In Proceedings of the 7th Python in Science Conference. 11--15.Google Scholar
- Will Hamilton, Zhitao Ying, and Jure Leskovec. 2017. Inductive representation learning on large graphs. In Proceedings of the 31st International Conference on Neural Information Processing Systems. I. Guyon, U. V. Luxburg, S. Bengio, H. Wallach, R. Fergus, S. Vishwanathan, and R. Garnett (Eds.), ACM, 1024--1034. Retrieved from http://papers.nips.cc/paper/6703-inductive-representation-learning-on-large-graphs.pdf.Google Scholar
- Takanori Hayashi, Takuya Akiba, and Yuichi Yoshida. 2015. Fully dynamic betweenness centrality maintenance on massive networks. 9 (2015), 48--59. DOI:https://doi.org/10.14778/2850578.2850580Google ScholarDigital Library
- Loc Hoang, Matteo Pontecorvi, Roshan Dathathri, Gurbinder Gill, Bozhi You, Keshav Pingali, and Vijaya Ramachandran. 2019. A round-efficient distributed betweenness centrality algorithm. In Proceedings of the 24th Symposium on Principles and Practice of Parallel Programming. Association for Computing Machinery, 272--286. DOI:https://doi.org/10.1145/3293883.3295729Google ScholarDigital Library
- Weihua Hu, Bowen Liu, Joseph Gomes, Marinka Zitnik, Percy Liang, Vijay Pande, and Jure Leskovec. 2019. Pre-training graph neural networks. In Proceedings of the 24th Symposium on Principles and Practice of Parallel Programming. 272-286Google Scholar
- S. Jin, Z. Huang, Y. Chen, D. Chavarría-Miranda, J. Feo, and P. C. Wong. 2010. A novel application of parallel betweenness centrality to power grid contingency analysis. In Proceedings of the 2010 IEEE International Symposium on Parallel Distributed Processing. 1--7. DOI:https://doi.org/10.1109/IPDPS.2010.5470400Google Scholar
- Maliackal Poulo Joy, Amy Brock, Donald E. Ingber, and Sui Huang. 2005. High-betweenness proteins in the yeast protein interaction network. Journal of Biomedicine and Biotechnology 2005, 2 (2005), 96--103. DOI:https://doi.org/10.1155/JBB.2005.96Google ScholarCross Ref
- M. Kas, M. Wachs, K. M. Carley, and L. R. Carley. 2013. Incremental algorithm for updating betweenness centrality in dynamically growing networks. In Proceedings of the 2013 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining 33--40. DOI:https://doi.org/10.1109/ASONAM.2013.6785684Google ScholarDigital Library
- M. G. Kendall. 1938. A new measure of rank correlation. Biometrika 30, 1 (1938), 81--93. DOI:https://doi.org/10.1093/biomet/30.1-2.81Google ScholarCross Ref
- Thomas N. Kipf and Max Welling. 2016. Semi-supervised classification with graph convolutional networks. In International Conference on Learning Representation (ICLR'17).Google Scholar
- N. Kourtellis, G. De Francisci Morales, and F. Bonchi. 2016. Scalable online betweenness centrality in evolving graphs. In Proceedings of the 2016 IEEE 32nd International Conference on Data Engineering. 1580--1581. DOI:https://doi.org/10.1109/ICDE.2016.7498421Google Scholar
- Kwang Hee Lee and Myoung Ho Kim. 2020. Linearization of dependency and sampling for participation-based betweenness centrality in very large B-hypergraphs. ACM Transactions on Knowledge Discovery from Data 14, 3 (2020), 25:1–25:41. DOI:https://doi.org/10.1145/3375399Google Scholar
- Min-Joong Lee, Jungmin Lee, Jaimie Yejean Park, Ryan H. Choi, and Chin-Wan Chung. 2012. QUBE: A quick algorithm for updating betweenness centrality. In Proceedings of the 21st international conference on World Wide Web. 351--360. DOI:https://doi.org/10.1145/2187836.2187884Google ScholarDigital Library
- Yujia Li, Oriol Vinyals, Chris Dyer, Razvan Pascanu, and Peter W. Battaglia. 2018. Learning deep generative models of graphs. In International Conference on Machine Learning.Google Scholar
- Hao Liao, Manuel Sebastian Mariani, Matúŝ Medo, Yi-Cheng Zhang, and Ming-Yang Zhou. 2017. Ranking in evolving complex networks. Physics Reports 689 (2017), 1--54. DOI:https://doi.org/10.1016/j.physrep.2017.05.001Google ScholarCross Ref
- Andreas Loukas. 2019. What graph neural networks cannot learn: Depth vs width. In Proceedings of the International Conference on Learning Representations.Google Scholar
- Haggai Maron, Heli Ben-Hamu, Hadar Serviansky, and Yaron Lipman. 2020. Provably powerful graph networks. In Advances in Neural Processing Systems 32 (NeurIPS'19).Google Scholar
- Sunil K. Maurya, Xin Liu, and Tsuyoshi Murata. 2019. Fast approximations of betweenness centrality using graph neural networks. In Proceedings of the 28th ACM International Conference on Information and Knowledge Management. ACM, 2149--2152.Google Scholar
- Matus Medo. 2013. Network-based information filtering algorithms: Ranking and recommendation. In Dynamics on and of Complex Networks, Volume 2: Applications to Time-Varying Dynamical Systems, Animesh Mukherjee, Monojit Choudhury, Fernando Peruani, Niloy Ganguly, and Bivas Mitra (Eds.). Springer, New York, 315--334. DOI:https://doi.org/10.1007/978-1-4614-6729-8_16Google Scholar
- Peter Mika. 2004. Social networks and the semantic web. In Proceedings of the 2004 IEEE/WIC/ACM International Conference on Web Intelligence. IEEE Computer Society, 285--291. Retrieved from http://dl.acm.org/citation.cfm?id=1025132.1026332.Google ScholarDigital Library
- Christopher Morris, Martin Ritzert, Matthias Fey, William L. Hamilton, Jan Eric Lenssen, Gaurav Rattan, and Martin Grohe. 2018. Weisfeiler and leman go neural: Higher-order graph neural networks. In Proceedings of the 32nd AAAI Conference on Artificial Intelligence. DOI:https://doi.org/10.1609/aaai.v33i01.33014602Google Scholar
- M. E. J. Newman. 2003. A measure of betweenness centrality based on random walks. Social Networks 27, 1 (2003), 39--54. DOI:https://doi.org/10.1016/j.socnet.2004.11.009Google ScholarCross Ref
- Kazuya Okamoto, Wei Chen, and Xiang-Yang Li. 2008. Ranking of closeness centrality for large-scale social networks. In Frontiers in Algorithmics, Franco P. Preparata, Xiaodong Wu, and Jianping Yin (Eds.). Lecture Notes in Computer Science, Springer, Berlin, 186--195.Google ScholarCross Ref
- Sarah Parisot, Sofia Ira Ktena, Enzo Ferrante, Matthew Lee, Ricardo Guerrerro Moreno, Ben Glocker, and Daniel Rueckert. 2017. Spectral graph convolutions for population-based disease prediction. In Proceedings of the International Conference on Medical Image Computing and Computer-Assisted Intervention, Maxime Descoteaux, Lena Maier-Hein, Alfred Franz, Pierre Jannin, D. Louis Collins, and Simon Duchesne (Eds.). Lecture Notes in Computer Science, Springer International Publishing, 177--185. DOI:https://doi.org/10.1007/978-3-319-66179-7_21Google Scholar
- Adam Paszke, Sam Gross, Soumith Chintala, Gregory Chanan, Edward Yang, Zachary DeVito, Zeming Lin, Alban Desmaison, Luca Antiga, and Adam Lerer. 2017. Automatic differentiation in PyTorch. In Proceedings ofthe31st Conference on Neural Information Processing Systems.Google Scholar
- Sen Pei, Lev Muchnik, José S. Andrade, Jr, Zhiming Zheng, and Hernán A. Makse. 2014. Searching for superspreaders of information in real-world social media. Scientific Reports 4, Article 5547 (2014). DOI:https://doi.org/10.1038/srep05547Google Scholar
- M. C. Pham and R. Klamma. 2010. The structure of the computer science knowledge network. In Proceedings of the 2010 International Conference on Advances in Social Networks Analysis and Mining. 17--24. DOI:https://doi.org/10.1109/ASONAM.2010.58Google ScholarDigital Library
- Rami Puzis, Yaniv Altshuler, Yuval Elovici, Shlomo Bekhor, Yoram Shiftan, and Alex Pentland. 2013. Augmented betweenness centrality for environmentally aware traffic monitoring in transportation networks. Journal of Intelligent Transportation Systems 17, 1 (2013), 91--105. DOI:https://doi.org/10.1080/15472450.2012.716663Google ScholarCross Ref
- Zarina Rakhimberdina and Tsuyoshi Murata. 2020. Linear graph convolutional model for diagnosing brain disorders. In Complex Networks and Their Applications VIII, Hocine Cherifi, Sabrina Gaito, José Fernendo Mendes, Esteban Moro, and Luis Mateus Rocha (Eds.). Studies in Computational Intelligence, Springer International Publishing, 815--826. DOI:https://doi.org/10.1007/978-3-030-36683-4_65Google Scholar
- Matteo Riondato and Evgenios M. Kornaropoulos. 2016. Fast approximation of betweenness centrality through sampling. Data Mining and Knowledge Discovery 30, 2 (2016), 438--475. DOI:https://doi.org/10.1007/s10618-015-0423-0Google ScholarDigital Library
- Gert Sabidussi. 1966. The centrality index of a graph. Psychometrika 31, 4 (1966), 581--603. DOI:https://doi.org/10.1007/BF02289527Google ScholarCross Ref
- Christian L. Staudt, Aleksejs Sazonovs, and Henning Meyerhenke. 2016. NetworKit: A tool suite for large-scale complex network analysis. Network Science 4, 4 (2016), 508--530. DOI:https://doi.org/10.1017/nws.2016.20Google ScholarCross Ref
- Petar Veliĉković, Guillem Cucurull, Arantxa Casanova, Adriana Romero, Pietro Liò, and Yoshua Bengio. 2018. Graph attention networks. In International Conference on Learning Representation (ICLR'18).Google Scholar
- Petar Veliĉković, William Fedus, William L. Hamilton, Pietro Liò, Yoshua Bengio, and R. Devon Hjelm. 2018. Deep graph infomax. In International Conference on Learning Representation (ICLR'19).Google Scholar
- Zhen Wang, Chris T. Bauch, Samit Bhattacharyya, Alberto d’Onofrio, Piero Manfredi, Matjaz Perc, Nicola Perra, Marcel Salathé, and Dawei Zhao. 2016. Statistical physics of vaccination. Physics Reports 664 (2016), 1--113. DOI:https://doi.org/10.1016/j.physrep.2016.10.006Google ScholarCross Ref
- Felix Wu, Amauri H. Souza, Tianyi Zhang, Christopher Fifty, Tao Yu, and Kilian Q. Weinberger. 2019. Simplifying graph convolutional networks. In Proceedings of the 36th International Conference on Machine Learning. 6861--6871.Google Scholar
- Zonghan Wu, Shirui Pan, Fengwen Chen, Guodong Long, Chengqi Zhang, and Philip S. Yu. 2021. A comprehensive survey on graph neural networks. IEEE Transactions on Neural Networks and Learning Systems 32, 1 (2021), 4--24.Google ScholarCross Ref
- Keyulu Xu, Weihua Hu, Jure Leskovec, and Stefanie Jegelka. 2018. How powerful are graph neural networks? In International Conference on Learning Representation (ICLR'19).Google Scholar
- Rex Ying, Ruining He, Kaifeng Chen, Pong Eksombatchai, William L. Hamilton, and Jure Leskovec. 2018. Graph convolutional neural networks for web-scale recommender systems. In Proceedings of the 24th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining. DOI:https://doi.org/10.1145/3219819.3219890Google ScholarDigital Library
- Rex Ying, Jiaxuan You, Christopher Morris, Xiang Ren, William L. Hamilton, and Jure Leskovec. 2018. Hierarchical graph representation learning with differentiable pooling. In Proceedings of the 32nd International Conference on Neural Information Processing Systems. Curran Associates Inc. 4805--4815. Retrieved from http://dl.acm.org/citation.cfm?id=3327345.3327389.Google Scholar
- Muhan Zhang, Zhicheng Cui, Marion Neumann, and Yixin Chen. 2018. An end-to-end deep learning architecture for graph classification. In Proceedings of the 32nd AAAI Conference on Artificial Intelligence. 4434--4445.Google Scholar
- Jie Zhou, Ganqu Cui, Zhengyan Zhang, Cheng Yang, Zhiyuan Liu, Lifeng Wang, Changcheng Li, and Maosong Sun. 2020. Graph neural networks: A review of methods and applications. AI Open 1 (2020), 57--81. https://www.sciencedirect.com/science/article/pii/S2666651021000012.Google ScholarCross Ref
Index Terms
- Graph Neural Networks for Fast Node Ranking Approximation
Recommendations
Fast Approximations of Betweenness Centrality with Graph Neural Networks
CIKM '19: Proceedings of the 28th ACM International Conference on Information and Knowledge ManagementBetweenness centrality is an important measure to find out influential nodes in networks in terms of information spread and connectivity. However, the exact calculation of betweenness centrality is computationally expensive. Although researchers have ...
Laplacian centrality: A new centrality measure for weighted networks
The centrality of vertices has been a key issue in network analysis. For unweighted networks where edges are just present or absent and have no weight attached, many centrality measures have been presented, such as degree, betweenness, closeness, ...
Graph Manipulations for Fast Centrality Computation
The betweenness and closeness metrics are widely used metrics in many network analysis applications. Yet, they are expensive to compute. For that reason, making the betweenness and closeness centrality computations faster is an important and well-...
Comments