HIDDEN MARKOV MODEL FOR DETECTION AND PREVENTION OF SQL INJECTION ATTACKS

Abstract

Web application have become un avoidable part in day to day life. Online activities such as banking, ecommerce, social networking etc., and these application stored in a back end of the database. It is very sensitive and confidential data, as because it might be get attacked. So security of Web database becomes an important thing. Most of the existing system have drawback such as it take more resources and times, reduce the efficiency of application, cannot generate test case dynamically. The proposing a novel method to detect malicious queries and handles it using HMM. For this, gathering a set of queries in order to compare  the incoming queries whether it is a injected or a genuine query. If it is injected one it will rejected or if it’s a genuine one it will allowed to perform actions. This technique can be easily ported to other languages and database platforms without requiring major modification.