BNADS: Big data Node Anomaly Detection in Social Networks

Abstract

In the recent years, detection of anomalies in social networks gains extensive attention to identify and classify irregular activities of online social users. With the popularity of online social networks, different types of anomalous activities are performed by anomalous users such as bullying, terrorist attack planning, and fraud information dissemination. In social networks, anomalous activities also called as abnormal behaviors compared to others in the same community structure. It is critical importance to identify and control such kind of anomalies or exceptions or outliers or illegal activities in social networks. We proposed elegant mechanism to identify anomalous nodes in static social network data. The work presented in this paper  is tested and implemented in big data environment using open source Hadoop tools such as HWX, Apache Spark etc. The suggested work gives the better performance results compared to the existing approaches.