Optimized Clustering for Anomaly Intrusion Detection

Abstract

Although conventional clustering algorithms have been used to classify data objects in a data set into the groups of similar data objects based on data similarity, they can be employed to extract the common knowledge i.e. properties of similar data objects commonly appearing in a set of transactions. The common knowledge of the activities in the transactions of a user is represented by the occurrence frequency of similar activities by the unit of a transaction as well as the repetitive ratio of similar activities in each transaction. This paper proposes an optimized clustering method for modeling the normal pattern of a user’s activities. Furthermore, it also addresses how to determine the optimal values of clustering parameters for a user as well as how to maintain identified common knowledge as a concise profile. As a result, it can be used to detect any anomalous behavior in an online transaction of the user.