Consensus ranking from a utility view

Consensus ranking problems have attracted much attention in the management science and data mining, where a group of k decision makers (DMs) is asked to rank n alternatives. Several different approaches have been suggested to combine the responses of DMs into a compromise or consensus ranking from a diverse perspective. However, they failed to work out in all situations. In this paper, we propose a new framework for consensus ranking from a utility view. Under this framework, we convert it to an optimization problem, which is proved to be NP-hard. To handle this challenging problem, three algorithms are put forward. Experiments show that Graph-based Ranking (GR) not only has the advantage of stability and validity on data of different scales, but also takes less time, which is quite suitable for the large scale consensus ranking.