Application of Soft Computing Techniques in Database Intrusion Detection

Abstract

newline With the rapid development of the World Wide Web and E-commerce, a concern of security is a very sensitive issue in this modern era of information and communication technology. A very large amount of time and effort has been invested by the research community working on database security to achieve high assurance of security and privacy. An important component of a secure database system is Intrusion Detection System which has the ability to successfully detect anomalous behaviour caused by applications and users. However, modeling the normal behaviour of a large number of users in a huge organization is quite tedious. Traditional techniques used in database security are not able to cope with the dynamic nature of the attacker. The use of soft computing approaches in database intrusion detection is an appealing concept due to its robustness and low solution cost and better rapport with reality. As part of the funded effort in database security, soft computing has proven to be capable of creating a system capable of detecting and characterizing anomalous behaviour which is composed of evolutionary computing tools with artificial neural networks and/or fuzzy logic. Artificial Neural Network (ANN) mimics the human brain that is composed of completely connected neurons and can have the ability to learn, categorize, and is able to predict the output. Genetic Algorithm (GA) is an optimization algorithm based on the principle of natural evolution. Soft Computing is a branch of artificial intelligence and ANN, Fuzzy logic and GA are the different dimensions of soft computing. In the current research, we have applied neural network and genetic algorithm is soft computing, as they aid in computation from unclear, incomplete input, and in the real world scenario, the output is not 100% accurate. This research focuses on building database intrusion detection models that can deal with the dynamic attitude of accessing the database by the users and able to detect intrusions with high accuracy and low false al