Conflicting information integration for decision support

The integration of information from multiple sources is a topic of increasing importance to the design of advanced decision support systems for a variety of situation assessment tasks. The integration of information is especially difficult when the multiple sources provide conflicting information. In general, approaches to this problem have used either probability-based techniques or techniques based on one of the newer theories of uncertainty. This paper surveys these current approaches to the conflicting information problem and then describes a new approach based on information theory.