Association Rules for Buffer Overflow Vulnerability Detection Using Machine Learning

Buffer overflow (BoF) vulnerability can be exploited by cyber-attackers to gain access and take control of information systems. As a result, effective vulnerability detection has become critical for information security in systems development. Using machine learning techniques, this study proposes a method for effective BoF vulnerability detection based on feature selection algorithms and association rule mining. The proposed method shows a significant enhancement of detection accuracy with 64% and 89% improvement, respectively, for two predefined BoF intrusion scenarios.