Argumentation-Based Query Answering under Uncertainty with Application to Cybersecurity

Decision support tools are key components of intelligent sociotechnical systems, and their successful implementation faces a variety challenges, including the multiplicity information sources, heterogeneous format, constant changes. Handling such challenges requires ability to analyze process inconsistent incomplete with varying degrees associated uncertainty. Moreover, some domains require system’s outputs be explainable interpretable; an example this is cyberthreat analysis (CTA) in cybersecurity domains. In paper, we first present P-DAQAP system, extension recently developed query-answering platform based on defeasible logic programming (DeLP) that incorporates probabilistic model focuses delivering these capabilities. After discussing details its design implementation, describing how it can applied CTA use case, report results empirical evaluation designed explore effectiveness efficiency possible world sampling-based approximate query answering approach addresses intractability exact computations.