224 A method to reject conflicting information for multiple information integration
Terumoto Komori, Yoshihiko Nomura, Hirokazu MATSUI, Norihiko Kato · The Proceedings of the Conference on Information Intelligence and Precision Equipment IIP · 2001
Dempster-Shafer probability theory has various advantages in the uncertainty representation and in the multiple information integration. However, information-integration leads to meaningless results in cases that the information contains incorrect data, i. e., some pieces of information are conflicting with others. Therefore, it is necessary to reject the inconsistent information before integration. Defining a new measure of conflict, the authors propose a method to reject the conflicting information.