Output details

Knowing Who to Watch : Efficiently Identifying Subtle Attackers

<19>This extended journal paper was invited to the special edition following the publication of a paper commended at MIST2009 (Accumulating Evidence of Insider Attacks, Chivers et al). The paper provides a novel solution to the problem of dealing with attack detection in very large event datasets: a Bayesian approach is extended with a method that provides distributed normalisation, and hence the ability to fully distribute the intrusion computations. The paper has prompted ongoing work into applications, and variants by other researchers (e.g. Sensing for Suspicion at Scale, Kalutarage et al).