Comparing rank-inducing scoring systems

Several methods for comparing rankings are available. These rankings are generally performed on the basis of some scores available for each item. We suggest that it may be better to compare the underlying scores themselves. To this effect, a metric has been provided which is equivalent to comparing the scores after fusing with another set of scores, making it theoretically interesting. This metric is a generalization of Kendall distance. In the present article, we assume the distribution of the scores to be fused with to be unknown. Some characteristics of the proposed methodology are studied and preliminary experimental results are reported