Computational intelligence in intrusion detection system for snort log using hadoop

Today Cyber-attacks are increasing and the security technologies so far we used are in capable of handling the large data sets of data. Now a day the attacks motive has changed from simple hacking, damaging to large scale network and server system. In the presence of fresh and formerly unknown activities, intrusion detection rate is inaccurate and low. Many research works are focused in this area. In all the security infrastructures, Network Intrusion Detection Systems (NIDS) have grown a standard and in separable component. For this reason, a new model has been proposed based on Big Data for detecting unknown attacks. The core idea of this paper is to analyze the activity of the network users, and to classify whether the user is normal or anomaly. For implementation of this project, we have used the snort, the tools which are used to capture the online behavior of network users. After collecting the network behavior, the dataset was analyzed with hadoop framework using C4.5 classification. The performance of the project was compared with KDD Dataset. The proposed approach improves the IDS performance and its capability is to provide quick response to various types of network attacks.