Data Mining For Web Security: UserWatcher *
Malika Mahoui, Mukesh Mohania · 2001
Data mining techniques have proved to be efficient for discovering interesting and useful patterns in large amount of data such as in Web documents. This paper investigates the use of mining techniques to secure Web access. We propose UserWatcher, a mining tool that integrates Web usage mining and Web content mining to find potential correlations between data that a user accesses and the data that he/she produces. Applied to security, UserWatcher can be used to detect information misusage by authorized users and to detect anomalies in current system policies.