Recognizing Intrusive Intention Based on Dynamic Bayesian Networks

Knowing an attackerpsilas intentions can significantly improve the effectiveness of a decision-making system. However, recognition such intentions and the attackerpsilas intended plans for achieving them is not an easy task because there are too many uncertain and dynamic factors in network environment. In this paper, intrusive intention recognition using dynamic Bayesian network is proposed to cope with uncertainty and dynamics in network security awareness. Furthermore, attack actions forecast based on goal recognition is given and discussed. Finally, feasibility and validity of this method are proved from the experiments.