Current studies to analyzing security protocols using formal methods require users to predefine authentication goals. Besides, they are unable to discover potential correlations between secure messages. This research attempts to analyze security protocols using data mining. This is done by extending the idea of association rule mining and converting the verification of protocols into computing the frequency and confidence of inconsistent secure messages. It provides a novel and efficient way to analyze security protocols and find out potential correlations between secure messages. The conducted experiments demonstrate our approaches.
Book Title : AI 2005: Advances in Artificial Intelligence
Field of Research
080199 Artificial Intelligence and Image Processing not elsewhere classified