Emerging Trends in ICT Security: Chapter 18. Real-time Network Intrusion Detection Using Hadoop-Based Bayesian Classifier (Emerging Trends in Computer Science and Applied Computing)

Over the years, many networks hosted by large companies or organizations have been crippled by intrusions launched with minimal effort. Such attacks have caused the loss of millions of dollars for the company and created serious security threats. As a result, network administrators and security experts across the globe have barricaded their networks with expensive Intrusion Detection Systems (IDS) to detect and take action in dealing with various network attacks. There is still a very challenging task to develop a cost-effective approach that can deal with network intrusions. Furthermore, large networks generate huge traffic data that serve as inputs for IDSes. In this chapter, we present a Network Intrusion Detection System (NIDS) built using Apache Hadoop and HStreaming, which can detect and alert administrators in real time. The system makes use of a simple yet versatile Naive Bayes classifier for predicting an attack. The experimental results show some promising outcomes.