A Joint Information Model for N-Best Ranking

In this paper, we present a method for modeling joint information when generating n-best lists. We apply the method to a novel task of characterizing the similarity of a group of terms where only a small set of many possible semantic properties may be displayed to a user. We demonstrate that considering the results jointly, by accounting for the information overlap between results, generates better n-best lists than considering them independently. We propose an information theoretic objective function for modeling the joint information in an n-best list and show empirical evidence that humans prefer the result sets produced by our joint model. Our results show with 95% confidence that the n-best lists generated by our joint ranking model are significantly different from a baseline independent model 50.0% ± 3.1% of the time, out of which they are preferred 76.6% ± 5.2% of the time.