A Selective Sampling Strategy for Label Ranking - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2006

A Selective Sampling Strategy for Label Ranking

Nicolas Usunier
  • Fonction : Auteur
  • PersonId : 933831
François Laviolette
  • Fonction : Auteur
Alexandre Lacasse
  • Fonction : Auteur

Résumé

We propose a novel active learning strategy based on the compression framework of [9] for label ranking functions which, given an input instance, predict a total order over a predefined set of alternatives. Our approach is theoretically motivated by an extension to ranking and active learning of Kääriäinen’s generalization bounds using unlabeled data [7], initially developed in the context of classification. The bounds we obtain suggest a selective sampling strategy provided that a sufficiently, yet reasonably large initial labeled dataset is provided. Experiments on Information Retrieval corpora from automatic text summarization and question/answering show that the proposed approach allows to substantially reduce the labeling effort in comparison to random and heuristic-based sampling strategies.

Dates et versions

hal-01337085 , version 1 (24-06-2016)

Identifiants

Citer

Massih-Reza Amini, Nicolas Usunier, François Laviolette, Alexandre Lacasse, Patrick Gallinari. A Selective Sampling Strategy for Label Ranking. European Conference on Machine Learning (ECML'06), Sep 2006, Berlin, Germany. pp.18-29, ⟨10.1007/11871842_7⟩. ⟨hal-01337085⟩
36 Consultations
0 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More