Efficient algorithms for ranking with SVMs

RankSVM (Herbrich et al, 2000; Joachims, 2002) is a pairwise method for designing ranking models. SVMLight is the only publicly available software for RankSVM. It is slow and, due to incomplete training with it, previous evaluations show RankSVM to have inferior ranking performance. We propose new methods based on primal Newton method to speed up RankSVM training and show that they are 5 orders of magnitude faster than SVMLight. Evaluation on the Letor benchmark datasets after complete training using such methods shows that the performance of RankSVM is excellent. Keywords ranking, support vector machines, AUC optimization

• Type: Journal
• Year: 2010
• Where: IR
• Authors: Olivier Chapelle, S. Sathiya Keerthi
• Keywords: Herbrich Et Al, IR 2010, Letor Benchmark Datasets, Natural Language Processing, Ranking