Adaptive context assessment and context management

Abstract

Adaptive context assessment and context management (CACM) methods opportunistically exploit non-traditional data sources to improve the robustness of information fusion systems. Adaptive CACM methods find relevant data in external data sources and create and refine predictive situational models based on the relevance, quality, and means of employing such data. These CACM methods also measure the conformity of this non-traditional data with Level 1-4 fusion system products. The method proposed here is developed as an extension to the Data Fusion and Resource Management (DF&RM) Dual Node Network (DNN) technical architecture by incorporating the CACM into the DNN fusion Level 4. Techniques are described that automatically learn to characterize and search non-traditional contextual data to enable fusion or comparison of data with organic data fusion systems products and ontologies. Non-traditional data can improve the quantity, quality, availability, timeliness, and diversity of the baseline fusion system sources and therefore can improve prediction, estimation accuracy and robustness at all levels of fusion.