Automated high-level reasoning for deception detection: Two scenarios demonstrated

Work that deals with automated approaches to identify deception is of great interest to the Department of Homeland Security, and the Department of Defense. The approach followed here is novel in that deception is identified on the basis of actions. A formalized logic programming approach to this problem is applied successfully to two dramatically different scenarios. What is fascinating is that the same reasoning architecture works successfully with both scenarios. Different, non-overlapping strategies have been discovered that assist in endowing the system with commonsense reasoning.