0 citations0 references

Order SVM: a kernel method for order learning based on generalized order statistics

Systems and Computers in Japan2004Vol. 36(1), pp. 35–43

Citations Over TimeTop 13% of 2004 papers

Hideto Kazawa, Tsutomu Hirao, Eisaku Maeda

Abstract

In this paper, we address the problem of learning the order of a group of items. This study assumes that items have unobserved “scores” and are ranked by the score. The goal is to derive an approximate scoring function given ordered lists of items without their scores. We extend the notion of order statistics to any order and prove that the distribution functions of the extended statistics are scoring functions. We then propose a learning algorithm, Order SVM, that can approximate the distribution functions. Order SVM is experimentally compared with other learning algorithms using artificial data and important sentence selection data. The results show that Order SVM performs well on both datasets whereas the other methods show good performance on only one of the datasets. © 2004 Wiley Periodicals, Inc. Syst Comp Jpn, 36(1): 35–43, 2005; Published online in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/scj.10630

Related Papers

→ Optimizing search engines using clickthrough data(2002)3,906 cited
→ Support vector learning for ordinal regression(1999)459 cited
→ 10.1162/1532443041827916(2000)393 cited
Learning Preference Relations for Information Retrieval(1998)