Text Categorization Techniques for Intrusion Detection Using Artificial Neural Network

There is a need to protect connected systems inorder to secure high data transmitted through internet. Intrusion Detection Systems (IDSs) are the latest technology used for this purpose. Although the field of IDSs is still developing, the systems that do exist are still not complete, in the sense that they are not able to detect all types of intrusions. Some attacks which are detected by various tools available today cannot be detected by other products, depending on the types and methods that they are built on. In this work, an artificial neural network using back-propagation algorithm has been used to implement the IDS with the kNN classifier, the frequencies of system calls are used to describe the program behaviour. Text categorization techniques are adopted to convert each process to a vector and calculate the similarities between two program activities. The k-Nearest Neighbour (kNN) classifier is used to classify program behaviour as normal or intrusive. Inspite of much related work had been done, this paper elucidates the implementation aspects of BPA for a real_time IDS. Thousand packet of information both normal and intrusion have been considered for implementation. The result of ID is very close to 99%. The topology of the ANN is (41 X 10 X 1). The network is converged with 550 iterations. Very huge amount of packets are to be evaluated to know the complete performance of the developed system.