Ranking Aggregation Based on Belief Function

In this work we consider the case of the ranking aggregation problem that includes the true ranking in its formulation. The goal is to find an estimation of an unknown true ranking given a set of rankings provided by different quality experts. This is the case when bioinformatic experts provide ranked items involved in an unknown biological phenomenon regulated by its own physical reality. We devise an innovative solution called Belief Ranking Estimator (BRE), based on the belief function framework that permits to represent beliefs on the correctness of each item rank as well as uncertainty on the quality of the rankings from the subjective point of view of the expert. Moreover, weights computed using a true-ranking estimator are applied to the original belief basic assignment in order to take into account the quality of the input rankings. The results of an empirical comparison of BRE with weighting schema against competitor methods for ranking aggregation show that our method improve...