Earticle

Any malicious attack on the database systems performed by an entrusted group of people having authorized access is called database insider attack. Even though the insider attack has been lively researched, it is still far from the practical application. We propose a new approach to improve the limitations in the previous researches. This approach has four objectives: (1) Multi-Preprocessing Algorithms to observe a query in multiple perspectives (2) Query Probabilities Based Database Insider Monitoring Methodology based upon Markov Mathematical Model to record insider’s behavioral patterns with query and query transition probabilities (3) Query Probabilities Time Series Graph to create metrics to monitor the insider’s behavior in order to predict insider attack, and (4) Multi-Criteria Query Probabilities Based Insider Attack Monitoring System containing the (1)–(3). The results from the evaluation show that the proposed system overcomes the limitations and is also capable to monitor insider’s behavioral data while the database is being updated.