Computational creativity for counterdeception in information fusion

Setting out from a discussion on why information fusion systems are vulnerable to deception, we note that exposing deception requires creativity. The burgeoning field of computational creativity is briefly reviewed and we sketch a high-level architecture for enhancing information fusion systems with computational creativity agents for the purpose of increasing robustness against deception. A more in depth exposition of how computational creativity integrates with a situation assessment sub-system based on statistical relational learning is provided and we note similarities between the different modes of this configuration and common human creativity techniques. One of these techniques is studied in a detailed example and used for revealing a suitable technical format for the creative input. Furthermore, we study how creative inputs are generated based on conceptual blending analysis applied to a library of deception stratagems.