验等多种统计工具的支撑下，我们发现共同好友的数量并不会对用户之间的轨迹相似性产生积极作用，反而是那些共同好友的多样性更强的用户有着更高的轨迹相似性（具体分析方法请见文献[6]）。也就是说，共同好友的来源多样性在影响好友之间的轨迹相似性上有着比数量更为重要的作用。我们进一步使用其它相似性指标重复实验，验证了结果的鲁棒性。
我们的研究不仅建立了人类行为相关性的分析框架，而且对于寻找行为相似性更高的用户进行轨迹预测[7]、商品推荐[8]、链路预测[9]等场景都有着直接或潜在的应用价值。
参考文献：
[1] Cho, E., Myers, S. A.& Leskovec, J. Friendship and mobility: User movement in location-basedsocial networks. In Proceedings of the 17th ACM SIGKDD International Conferenceon Knowledge Discovery and Data Mining, 1082–1090 (ACM, 2011).
[2] Liben-Nowell, D., Novak,J., Kumar, R., Raghavan, P. & Tomkins, A. Geographic routing in socialnetworks. Proc. Natl. Acad. Sci. USA 102, 11623–11628 (2005).
[3] Toole, J. L., Herrera-Yaqüe,C., Schneider, C. M. & González, M. C. Coupling human mobility and socialties. J. R. Soc. Interface 12, 20141128 (2015).
[4] Wang, D., Pedreschi, D.,Song, C., Giannotti, F. & Barabási, A.-L. Human mobility, social ties, andlink prediction. In Proceedings of the 17th ACM SIGKDD International Conferenceon Knowledge Discovery and Data Mining, 1100–1108 (ACM, 2011).
[5] Ugander, J., Backstrom,L., Marlow, C. & Kleinberg, J. Structural diversity in social contagion.Proc. Natl. Acad. Sci. USA 109, 5962–5966 (2012).
[6] Fan, C. et al. Correlationbetween social proximity and mobility similarity. Scientific Reports 7, 11975(2017).
[7] Lian, D. et al. MiningLocation-Based Social Networks: A Predictive Perspective. IEEE Data EngineeringBulletin 38, 35–46 (2015).
[8] Lü, L. et al. Recommendersystems. Physics Reports 519, 1–49 (2012).
[9] Lü, L. & Zhou, T. Linkprediction in complex networks: A survey. Physica A: Statistical Mechanics andits Applications 390, 1150–1170 (2011).
论文信息：
Chao Fan, Yiding Liu, Junming Huang, Zhihai Rong, TaoZhou. Correlation between social proximity and mobility similarity. Scientific Reports, 2017, 7: 11975.
论文链接（可免费下载）：
https://www.nature.com/articles/s41598-017-12274-x
| 欢迎光临 社会网络分析论坛 social network analysis forum (http://www.snachina.com/) | Powered by Discuz! X3.3 |