A Dempster-Shafer theoretic conditional approach to evidence updating for fusion of hard and soft data

Fusion of hard data with soft data is an issue that has attracted recent attention. An effective fusion strategy requires an analytical framework that can capture the uncertainty inherent in hard and soft data. For instance, computational linguistic parsing of text-based data generates logical propositions that inherently possess significant semantic ambiguity. An effective fusion framework must exploit the respective advantages of hard and soft data while mitigating their particular weaknesses. In this paper, we describe a Dempster-Shafer theoretic approach to hard and soft data fusion that relies upon the novel conditional approach to updating. The conditional approach engenders a more flexible method that allows for tuning and adapting update strategies. When computational complexity concerns are taken into account, it also provides guidance on how evidence could be ordered for updating. This has important implications in working with models that convert propositional logic statements from text into Dempster-Shafer theoretic form.