The rank aggregation problem has been studied extensively in recentyears with a focus on how to combine several different rankers to obtain a consensus aggregate ranker. We study the rank aggregation problem from a different perspective: how the individual input rankers impact the performance of the aggregate ranker. We develop a general statistical framework based on a model of how the individual rankers depend on the ground truth ranker. Within this framework, we study the performance of our novel aggregators together with other well known methods. Our results show that the relative performance of aggregators varies considerably depending on how the input rankers relate to the ground truth in terms on noise and misinformation. We develop two measures, cluster quality and rank variance, to measure misinformation and noise of a given set of ranked lists. Further, we develop a cost based decision method to find the least risky class for a new set of ranked lists and show that this decision method outperforms any static rank aggregation method by thorough experimentation.

Bio:Sibel Adali is an Associate Professor of Computer Science at Rensselaer Polytechnic Institute (RPI). She has received her B.S. from Bilkent University and M.S. and Ph.D. from University of Maryland. She has been a faculty at RPI since 1996. Her Ph.D. thesis concentrated on the integration of information from multiple heterogeneous information sources. Her current research focuses on information retrieval, ranking and query processing methods for multimedia and heterogeneous information systems. Some of the most current research topics include scientific data applications,extracting meaning from user actions and managing the data spaces for personal information management systems. She will present some of this future work and give a brief introduction to the research being conducted at RPI.