论文部分内容阅读
在线社交网络中的Sybil账号日益猖獗,他们会制造各种恶意活动,这些严重危害到了社交网络和用户的安全.针对Sybil账号检测这个问题,提出一个非常高效的Sybil账号检测模型.该模型提出使用受害者预测来提高检测准确性,将抽取的特征属性进行建模得到分类器,使用分类器进行受害者预测.再将预测结果应用到社交网络图模型中,最后,使用修改的随机游走对图节点进行排序.实验结果证明Sybil账号节点排在序列的底部,从而将Sybil账号从正常账号中分离出来.该模型的准确性达到了95%,这表明本文提出的检测模型是可行和有效的.
Sybil accounts in online social networks are becoming increasingly rampant, they will create a variety of malicious activities, which seriously endanger the safety of social networks and users.To detect this problem Sybil account, put forward a very efficient Sybil account detection model.This model proposed the use of The victim prediction is used to improve the detection accuracy, the extracted feature attributes are modeled to obtain the classifier, the classifier is used to predict the victim, the prediction result is applied to the social network graph model, and finally, the modified random walk The results show that the Sybil account node is at the bottom of the sequence, so that the Sybil account is separated from the normal account.The accuracy of the model reaches 95%, indicating that the proposed detection model is feasible and effective .