(2) 评估:图模型相对监督模型如分类器,评估难度很大,可能很难给出特别精准的评估效果,但是依然可以找到方法进行部分评估。评估分离线评估和在线评估。离线评估方法有 a) 交叉验证,评估历史一段时间坏样本的覆盖率,好样本的误伤率 b) 利用其它模型交互验证。在线评估: 主要是A/B Test ,设计线上评估指标,如登录 成功率,交易成功率,验证成功率等等,评估这些欺诈节点在在这些评估指标上的量化效果。
[2] Akoglu L, Tong H, Koutra D. Graph based anomaly detection and description: a survey[J]. Data Mining and Knowledge Discovery, 2015, 29(3): 626-688.
[3] Sensarma D, Sarma S S. A survey on different graph based anomaly detection techniques[J]. Indian Journal of Science and Technology, 2015, 8(31).
[4]
Aktas M S, Nacar M A, Menczer F. Personalizing pagerank based on domain profiles[C]//Proc. of WebKDD. 2004: 22-25.
[5] Krishnan V, Raj R. Web spam detection with anti-trust rank[C]//AIRWeb. 2006, 6: 37-40.
[6] Gyöngyi Z, Garcia-Molina H, Pedersen J. Combating web spam with trustrank[C]//Proceedings of the Thirtieth international conference on Very large data bases-Volume 30. VLDB Endowment, 2004: 576-587.
[7]
Jiang M, Cui P, Beutel A, et al. Catchsync: catching synchronized behavior in large directed graphs[C]//Proceedings of the 20th ACM SIGKDD international conference on Knowledge discovery and data mining. ACM, 2014: 941-950.
[8]
Raghavan U N, Albert R, Kumara S. Near linear time algorithm to detect community structures in large-scale networks[J]. Physical review E, 2007, 76(3): 036106.
[9]
Xie J, Szymanski B K. Labelrank: A stabilized label propagation algorithm for community detection in networks[C]//Network Science Workshop (NSW), 2013 IEEE 2nd. IEEE, 2013: 138-143.
[10]
Blondel V D, Guillaume J L, Lambiotte R, et al. Fast unfolding of communities in large networks[J]. Journal of statistical mechanics: theory and experiment, 2008, 2008(10): P10008.
[11]
Sun J, Qu H, Chakrabarti D, et al. Neighborhood formation and anomaly detection in bipartite graphs[C]//Data Mining, Fifth IEEE International Conference on. IEEE, 2005: 8 pp.