Abstract
Globally the internet is been accessed by enormous people within their restricted domains. When the client and server exchange messages among each other, there is an activity that can be observed in log files. Log files give a detailed description of the activities that occur in a network that shows the IP address, login and logout durations, the user's behavior etc. There are several types of attacks occurring from the internet. Our focus of research in this paper is Denial of Service (DoS) attacks with the help of pattern recognition techniques in data mining. Through which the Denial of Service attack is identified. Denial of service is a very dangerous attack that jeopardizes the IT resources of an organization by overloading with imitation messages or multiple requests from unauthorized users.