High-Level fusion for intelligence applications using Recombinant Cognition Synthesis

Intelligence applications exploit heterogeneous data using High-Level fusion systems to gain information superiority. Whereas Low-Level fusion systems have well established frameworks, High-Level fusion has not yet achieved the same level of maturity. Most High-Level systems implement specialized algorithms that yield useful results, albeit for a very narrow input space, and are characterized by stove-pipe architectures and a fragmented workflow. Recombinant Cognition Synthesis bridges the implementation gap of existing fusion models by defining a comprehensive framework of semantic, temporal, and geospatial enablers comprising the primitives, functions, and models, which through a recombinant workflow, maximize the data exploitation value-chain. This paper presents a methodology and the underlying architectural components necessary to implement a unified High-Level fusion intelligence application, followed by a case study that demonstrates the resulting improvements in knowledge discovery and predictive accuracy.

Research highlights
► A Trans-Dimensional framework facilitates domain space entity enrichment.
► Recombination unifies fusion models by leveraging an integrated Enterprise Architecture.
► A recombinant approach is the most effective way to exploit the trans-dimensional domain space.
► Fusion Metadata feedback loops are essential for optimizing Process Refinement.