Dos Attack and Suspicious Activity Detection for a Book Application

Abstract

Globally the internet is been accessed by several people within their restricted domains. When the client and server exchange message among each other, there is an activity that can be observed and tracked in detail of the activities that occur in a network that shows the, login and logout durations, the user’s behaviour etc. there are several types of attacks occurring from the internet. In this work the first focus is to provide product recommendations based on the collaborative filtering and content based recommendations. In personal setting the user in order to filter the products based on the transaction log maintained in terms of order information and order details. In Content based recommendations the recommendations related to the product are made based on the transactions performed by the end user. The Session tracking is performed for every click action and every navigation of the user and then behaviour based habitat file is generated. Two kinds of intrusion are detected one is behaviour change using Least Common Sub Square algorithm and then Dos Attack which is repeated action performed by the user within the limited time frame.