Merging operator selection in possibility theory

In data processing, merging of data requires definition of conflict degree between source information in order to select the appropriate merging operator. The diversity of merging operators and the variability of information precision and source reliability generate great difficulty in the selection of the appropriate merging operator. Possibility theory and Dempster Shafer (DS) theory are two main formalisms in modeling with uncertain information. The common aspect of these two theories is how to measure the degree of conflict. In this paper, we propose an original strategy to select a merging operator according an appropriate conflict degree investigation. We deploy different operator modalities proposed in possibility theory. We discuss how to select the operator modality based on the uncertainty degree developed in possibility theory and the distance between betting commitments developed in DS theory. The proposed strategy allows a traitor made merging of several information sources and to provide an informative possibility distribution.