基于交互行为的在线社会网络水军检测方法

Interaction based on method for spam detection in online social networks

网络水军对广告、谣言、木马和恶意链接进行传播,不仅干扰用户对在线社会网络的正常访问,还可能引发网络安全、社会稳定等方面的问题。针对网络水军信息传播的特点,提出基于交互行为的信息传播模型。模型根据不同传播主体间的交互定义特征来量化传播行为,使用决策树方法对水军传播的信息进行检测。通过新浪微博的真实数据分析传播模型并验证检测方法,结果表明检测方法能够对微博中水军信息进行有效检测。

In online social networks, advertisements, rumors and malicious links are propagated by spammers arbitrarily. They not only disturb users＇ usualaccess, but also bring about network security threats and social panics. In an attempt to deal with the spam problems, an information diffusion model was proposed to capture the features of spam propagation. Propagation behaviors are quantitatively analyzed to detect spam messages with a decision tree-based method. The effectiveness of proposed detection model is evaluated with real data from the micro-bloggingnetwork of Sina. The experimental results show that proposed model can effectively detect spams in Sina micro-bloggingnetwork.