BRNADS: Big data real-time node anomaly detection in social networks

Abstract

Anomalous activities are increased during recent years due to large usage of Social Networks such as Twitter, Facebook, etc. The processes of identifying the abnormal or illegal or irregular activities in the Social Networks are called anomaly detection. Many researchers are proposed different mechanisms to identify unlawful activities of the Social Networks users. It is very much important and useful to detect anomalous activities to prevent huge losses. We can carry out anomaly detection in either static or dynamic way. In this paper, we are proposing a dynamic anomaly detection mechanism to investigate the outlier's in the Social Networks. The proposed work takes the real time stream data using big data ecosystem called Apache Kafka and processing are done using spark streaming. The algorithm works on all types of data, implemented and tested in Big data environment using Java and Apache Spark.