摘要
为解决Folksonomy存在垃圾标签的问题,提出垃圾标签检测模型。利用向量空间模型表征用户特征,再用支持向量机将Folksonomy用户二分类。通过检测出隐藏在正常用户群体中的垃圾投放人,以此减少垃圾标签数量。实验结果表明,基于支持向量机的垃圾标签检测模型具有更高的分类精度,优于其他检测方法。
The popular social bookmarking sites were always attacked by social spam. This paper designed a SVM-based social spam detection model to solve this problem. That was using VSM to build the user model ,and then divided the users of the sites into two classes by SVM,of which one was the normal,the other was spammer. So cut off the social spam by reducing the spammer. The result of the experiment shows that the classification accuracy of SVM-based social spam detection model is higher than others.
出处
《计算机应用研究》
CSCD
北大核心
2010年第10期3893-3895,共3页
Application Research of Computers
关键词
垃圾标签
社会化标签系统
支持向量机
检测模型
social spam
social bookmark system
SVM（ support vector machines）
detection model
