An Efficient Method to Measure Evidence Conflict

Abstract

Abstract-Dempster Shafer evidence theory, as an uncertain information fusion technology, is widely used in various fields of information fusion. However, when there is a highly conflict between two pieces of evidence, counterintuitive results are obtained by classical Dempster's combination rule. Therefore, it is very important to measure the conflict between two pieces of evidence. Based on the analysis of some typical conflict measurement methods, a new model to represent conflict was constructed with the Euclidean distance function and the Jaccard similarity coefficient. Some numerical examples illustrate that the proposed method can measure the degree of conflict between the two pieces of evidence and overcome the shortcomings of the classical conflictive coefficient, Jousselme evidence distance and Pignistic probability distance to a certain extent.