Day by day more and more web based social media applications such as Face Book, Twitter, LinkedIn, etc. are established to improve the user’s performance. But whatever the security concerns faced by them are not exposed to the users, lot of user’s sensitive data might get hacked and thrown somewhere else. In this paper in order to safe guard the data which are gathered in social media, we are preventing the social media applications from Sql Injection, Denial of services (DDOS), Cross Browser Attack (XSS), Phishing, Cross Browser Request Forgery (CSRF), Click Jacking, Inference Attack, etc. We propose a probability model of the mentioning behavior of a social network user, and propose to detect the emergence of a new topic from the anomalies measured through the model. Aggressive unusual grades from hundreds of users, we show that we can deduction emerging topics only based on the reply/mention relationships in social network posts. We demonstrate our technique in several real data sets we gathered from Twitter. The check-up show that the suggest mention-anomaly-based access can deduction new topics at least as early as text-anomaly-based approaches, and in some cases much earlier when the topic is poorly identified by the textual contents in posts.