Dempster-Shafer Theory: combination of information using contextual knowledge

Abstract – The aim of this paper is to investigate how to improve the process of information combination, using the Dempster-Shafer Theory (DST). In presence of an overload of information and an unknown environ-ment, the reliability of the sources of information or the sensors is usually unknown and thus cannot be used to refine the fusion process. In a previous paper [1], the authors have investigated different techniques to evalu-ate contextual knowledge from a set of mass functions (membership of a BPA to a set of BPAs, relative relia-bilities of BPAs, credibility degrees, etc.). The purpose of this paper is to investigate how to use the contextual knowledge in order to improve the fusion process