Model Neutral Data Representation Using the Theory of Corobs

We propose to demonstrate a powerful, general approach, the Theory of Corobs, to address DARPA's identified need for canonical data structure and access mechanisms to support model interoperability. The technology we will produce under this contract will enable the construction of robust data warehousing supporting a wide range of "looks" of the data to the outside world, true generic knowledge bases. The Theory of Corobs is based on mathematical object known as a "corob," a strong candidate for the primary information processing token in living neural systems. The mathematics of corobs is rich and deep. We will demonstrate the relevant properties of corobs and show how to use them to implement model neutral data representations. We will build two models, a discrete event simulation and a Petri-net simulation and demonstrate that our corob system supports both, applying appropriate metrics. In Phase II, we will expand this work in cooperation with the OPR to develop tools that capture this technology for DARPA purposes and wider commercialization.