Half transductive ranking

Bing Bai, Jason Weston, David Grangier, Ronan Collobert, Corinna Cortes, Mehryar Mohri

Research output: Contribution to journalArticle

Abstract

We study the standard retrieval task of ranking a fixed set of items given a previously unseen query and pose it as the half transductive ranking problem. The task is transductive as the set of items is fixed. Transductive representations (where the vector representation of each example is learned) allow the generation of highly nonlinear embeddings that capture object relationships without relying on a specific choice of features, and require only relatively simple optimization. Unfortunately, they have no direct outof- sample extension. Inductive approaches on the other hand allow for the representation of unknown queries. We describe algorithms for this setting which have the advantages of both transductive and inductive approaches, and can be applied in unsupervised (either reconstruction-based or graph-based) and supervised ranking setups. We show empirically that our methods give strong performance on all three tasks.

Original languageEnglish (US)
Pages (from-to)49-56
Number of pages8
JournalJournal of Machine Learning Research
Volume9
StatePublished - 2010

Fingerprint

Ranking
Query
Retrieval
Unknown
Optimization
Graph in graph theory
Relationships
Object
Standards

ASJC Scopus subject areas

  • Artificial Intelligence
  • Software
  • Control and Systems Engineering
  • Statistics and Probability

Cite this

Bai, B., Weston, J., Grangier, D., Collobert, R., Cortes, C., & Mohri, M. (2010). Half transductive ranking. Journal of Machine Learning Research, 9, 49-56.

Half transductive ranking. / Bai, Bing; Weston, Jason; Grangier, David; Collobert, Ronan; Cortes, Corinna; Mohri, Mehryar.

In: Journal of Machine Learning Research, Vol. 9, 2010, p. 49-56.

Research output: Contribution to journalArticle

Bai, B, Weston, J, Grangier, D, Collobert, R, Cortes, C & Mohri, M 2010, 'Half transductive ranking', Journal of Machine Learning Research, vol. 9, pp. 49-56.
Bai B, Weston J, Grangier D, Collobert R, Cortes C, Mohri M. Half transductive ranking. Journal of Machine Learning Research. 2010;9:49-56.
Bai, Bing ; Weston, Jason ; Grangier, David ; Collobert, Ronan ; Cortes, Corinna ; Mohri, Mehryar. / Half transductive ranking. In: Journal of Machine Learning Research. 2010 ; Vol. 9. pp. 49-56.
@article{bb3c3c7c86a54acbab3f0e35def7a5f6,
title = "Half transductive ranking",
abstract = "We study the standard retrieval task of ranking a fixed set of items given a previously unseen query and pose it as the half transductive ranking problem. The task is transductive as the set of items is fixed. Transductive representations (where the vector representation of each example is learned) allow the generation of highly nonlinear embeddings that capture object relationships without relying on a specific choice of features, and require only relatively simple optimization. Unfortunately, they have no direct outof- sample extension. Inductive approaches on the other hand allow for the representation of unknown queries. We describe algorithms for this setting which have the advantages of both transductive and inductive approaches, and can be applied in unsupervised (either reconstruction-based or graph-based) and supervised ranking setups. We show empirically that our methods give strong performance on all three tasks.",
author = "Bing Bai and Jason Weston and David Grangier and Ronan Collobert and Corinna Cortes and Mehryar Mohri",
year = "2010",
language = "English (US)",
volume = "9",
pages = "49--56",
journal = "Journal of Machine Learning Research",
issn = "1532-4435",
publisher = "Microtome Publishing",

}

TY - JOUR

T1 - Half transductive ranking

AU - Bai, Bing

AU - Weston, Jason

AU - Grangier, David

AU - Collobert, Ronan

AU - Cortes, Corinna

AU - Mohri, Mehryar

PY - 2010

Y1 - 2010

N2 - We study the standard retrieval task of ranking a fixed set of items given a previously unseen query and pose it as the half transductive ranking problem. The task is transductive as the set of items is fixed. Transductive representations (where the vector representation of each example is learned) allow the generation of highly nonlinear embeddings that capture object relationships without relying on a specific choice of features, and require only relatively simple optimization. Unfortunately, they have no direct outof- sample extension. Inductive approaches on the other hand allow for the representation of unknown queries. We describe algorithms for this setting which have the advantages of both transductive and inductive approaches, and can be applied in unsupervised (either reconstruction-based or graph-based) and supervised ranking setups. We show empirically that our methods give strong performance on all three tasks.

AB - We study the standard retrieval task of ranking a fixed set of items given a previously unseen query and pose it as the half transductive ranking problem. The task is transductive as the set of items is fixed. Transductive representations (where the vector representation of each example is learned) allow the generation of highly nonlinear embeddings that capture object relationships without relying on a specific choice of features, and require only relatively simple optimization. Unfortunately, they have no direct outof- sample extension. Inductive approaches on the other hand allow for the representation of unknown queries. We describe algorithms for this setting which have the advantages of both transductive and inductive approaches, and can be applied in unsupervised (either reconstruction-based or graph-based) and supervised ranking setups. We show empirically that our methods give strong performance on all three tasks.

UR - http://www.scopus.com/inward/record.url?scp=84862290681&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84862290681&partnerID=8YFLogxK

M3 - Article

VL - 9

SP - 49

EP - 56

JO - Journal of Machine Learning Research

JF - Journal of Machine Learning Research

SN - 1532-4435

ER -