Ranking with Unlabeled Data: A First Study

Abstract : In this paper, we present a general learning framework which treats the ranking problem for various Information Retrieval tasks. We extend the training set generalization error bound proposed by Matti Kaariannan to the ranking case and show that the use of unlabeled data can be beneficial for learning a ranking function. We finally discuss open issues regarding the use of the unlabeled data during training a ranking function.
Document type :
Conference papers
Complete list of metadatas

Contributor : Lip6 Publications <>
Submitted on : Wednesday, March 15, 2017 - 2:25:00 PM
Last modification on : Thursday, March 21, 2019 - 1:04:52 PM


  • HAL Id : hal-01490506, version 1


Nicolas Usunier, Tuong Vinh Truong, Massih-Reza Amini, Patrick Gallinari. Ranking with Unlabeled Data: A First Study. NIPS'05 Workshop on Learning to Rank (NIPS'05-LR), Dec 2005, Whistler, BC, Canada. pp.24-28. ⟨hal-01490506⟩



Record views