Counter-deception in information fusion

摘要

In this article, we develop an entropy-based degree of falsity and combine it with a previously developed conflict-based degree of falsity in order to grade all belief functions. The new entropy-based degree of falsity is based on observing changes in entropy that are not consistent with combining only truthful information. With this measure, we can identify deliberately deceptive information and exclude it from the information fusion process. An experiment is performed comparing conflict and entropy measures and their combination. The effectiveness of the combination of the two measures is suggested. (C) 2017 Elsevier Inc. All rights reserved.