Yahoo! Learning to Rank Challenge 2011

6 papers
The 2011 Yahoo! Learning to Rank Challenge focused on advancing machine learning techniques for web search ranking, with a strong emphasis on ensemble methods and generalization. Research themes included the use of extremely randomized trees and calibrated AdaBoost to improve ranking accuracy. Notable contributions explored the application of gradient boosted regression trees (GBRT), such as the YetiRank algorithm, which was utilized to address the challenge's transfer learning track. A significant portion of the research also addressed transfer learning and model robustness. Scholars investigated methods for domain adaptation and the mitigation of negative transfer through feature decoupling, particularly in low-data regimes. Additional work focused on handling label noise through latent transition matrices and asymmetric noise models. The challenge also featured developments in differentiable ranking and sorting for gradient-based learning, alongside strategies for efficient ensemble inference. These efforts collectively aimed to enhance the efficiency and reliability of ranking systems in complex information retrieval environments.

Top Research Directions

The 18 research directions with Yahoo! Learning to Rank Challenge 2011 papers, ranked by paper count. Share is the fraction of the edition’s papers in the direction.

#DirectionPapersShare
1Differentiable Relaxations for Sorting and Ranking116.7%
2Efficient Document Scoring for Real-Time Neural Ranking116.7%
3Ensembling Decision Trees for Complex Non-linear Modeling116.7%
4Feature Whitening in Representation Learning116.7%
5Informed Initialization for Bayesian Optimization116.7%
6Learning Theory for Pairwise Data Dependencies116.7%
7Learning with Asymmetric and Class-Conditional Label Noise116.7%
8Memory Overhead in Multi-Model Inference Pipelines116.7%
9Mitigating Negative Transfer in Cross Domain Learning116.7%
10Noise Transition Matrix Estimation for Noisy Labels116.7%
11Parallelization Constraints in Sequential Ensemble Training116.7%
12Stabilizing Ranked Lists Against Data Perturbations116.7%
13Surrogate Regret Bounds for Learning to Rank116.7%
14The Critical Role of Systematic Hyperparameter Optimization116.7%
15The Limits of Ensemble Scaling116.7%
16Transfer Learning via Smooth Residual Offsets116.7%
17Unbiased Learning to Rank from Implicit Feedback116.7%
18Weighted Model Ensembling116.7%

Download all directions as CSV

Institutions

Institutions credited by Yahoo! Learning to Rank Challenge 2011 papers: a paper credits every institution named in its affiliations, once. 13 institutions share 6 credited papers.

#InstitutionPapersShare
1Yahoo (United States) 233.3%
2Centre Inria de Saclay 116.7%
3Centre National de la Recherche Scientifique 116.7%
4Hungarian Academy of Sciences 116.7%
5Institut national de recherche en sciences et technologies du numérique 116.7%
6Laboratoire de Recherche en Informatique 116.7%
7Microsoft (United States) 116.7%
8Technische Universität Darmstadt 116.7%
9University of Liège 116.7%
10University of Szeged 116.7%
11Université Paris-Sud 116.7%
12Washington University in St. Louis 116.7%
13Yandex (Russia) 116.7%

Download all institutions as CSV

About this data

Papers. Yahoo! Learning to Rank Challenge 2011 lists 6 papers on Lacuna, duplicate versions counted once; rejected and withdrawn submissions are left out.

Directions. Directions are Lacuna’s clustering of the concept elements extracted from each paper, so a paper can belong to several and direction counts add up to more than the paper count. A direction’s share is its papers divided by the edition’s papers.

Institutions. A paper credits each institution named in its affiliations, resolved to a ROR record, once. 6 of 6 papers (100.0%) credit at least one institution and are the base for institution shares; the rest have no resolved affiliation yet.

Data. Counts come from Lacuna’s data build of 2026-11-20 and change as papers are added. Tables as CSV: directions · institutions. Cite as: Lacuna, “Yahoo! Learning to Rank Challenge 2011”, https://lacuna.tiptreesystems.com/venue/yahoo-learning-to-rank-challenge/2011.

Top Authors

AuthorPapers
Yi Chang2
Andrey Gulin1
Balázs Kégl1
Gilles Louppe1
György Szarvas1
Kilian Q Weinberger1
Pierre Geurts1
Robert Istvan Busa-Fekete1
Tie-Yan Liu1

Impactful Papers

Papers from Yahoo! Learning to Rank Challenge 2011 that shaped research clusters.

Foundational

Influential

Papers

Title
Future directions in learning to rank
Learning to rank with extremely randomized trees
Ranking by calibrated AdaBoost
Web-Search Ranking with Initialized Gradient Boosted Regression Trees
Winning The Transfer Learning Track of Yahoo!'s Learning To Rank Challenge with YetiRank
Yahoo! Learning to Rank Challenge Overview