Applying Data Mining to Data Security

The decrease in the costs of storage devices and the advances in data warehousing advocate a new area of research knowledge discovery in databases. There have been research e orts that apply these techniques to computer security. Unlike previous e orts which focus on operating system and network layer, we use data mining techniques to detect misuse in database systems at the application layer where the database schema is taken into consideration. In this paper, an overview of data mining is presented. It is illustrated how data mining can be applied to computer security in building an inference engine and in misuse detection. A novel and generic framework for an anomaly detection system based on data mining is presented. Finally, the advantages and challenges of applying data mining techniques to misuse detection in database systems are discussed.