Search Results for author: Harrie Oosterhuis

Found 20 papers, 16 papers with code

Closing the Gender Wage Gap: Adversarial Fairness in Job Recommendation

1 code implementation20 Sep 2022 Clara Rus, Jeffrey Luppes, Harrie Oosterhuis, Gido H. Schoenmacker

We conclude that adversarial debiasing of word representations can increase real-world fairness of systems and thus may be part of the solution for creating fairness-aware recommendation systems.

Fairness Recommendation Systems

The Bandwagon Effect: Not Just Another Bias

1 code implementation25 Jun 2022 Norman Knyazev, Harrie Oosterhuis

Optimizing recommender systems based on user interaction data is mainly seen as a problem of dealing with selection bias, where most existing work assumes that interactions from different users are independent.

Recommendation Systems Selection bias

Reaching the End of Unbiasedness: Uncovering Implicit Limitations of Click-Based Learning to Rank

no code implementations24 Jun 2022 Harrie Oosterhuis

Thus, in contrast with limitations that follow from explicit assumptions, our aim is to recognize limitations that the field is currently unaware of.


State Encoders in Reinforcement Learning for Recommendation: A Reproducibility Study

1 code implementation10 May 2022 Jin Huang, Harrie Oosterhuis, Bunyamin Cetinkaya, Thijs Rood, Maarten de Rijke

In response to these shortcomings, we reproduce and expand on the existing comparison of attention-based state encoders (1) in the publicly available debiased RL4Rec SOFA simulator with (2) a different RL method, (3) more state encoders, and (4) a different dataset.


Learning-to-Rank at the Speed of Sampling: Plackett-Luce Gradient Estimation With Minimal Computational Complexity

1 code implementation22 Apr 2022 Harrie Oosterhuis

Plackett-Luce gradient estimation enables the optimization of stochastic ranking models within feasible time constraints through sampling techniques.


Doubly-Robust Estimation for Correcting Position-Bias in Click Feedback for Unbiased Learning to Rank

1 code implementation31 Mar 2022 Harrie Oosterhuis

The prevalent approach to unbiased click-based learning-to-rank (LTR) is based on counterfactual inverse-propensity-scoring (IPS) estimation.

General Reinforcement Learning Learning-To-Rank

It Is Different When Items Are Older: Debiasing Recommendations When Selection Bias and User Preferences Are Dynamic

1 code implementation24 Nov 2021 Jin Huang, Harrie Oosterhuis, Maarten de Rijke

We theoretically show that in a dynamic scenario in which both the selection bias and user preferences are dynamic, existing debiasing methods are no longer unbiased.

Recommendation Systems Selection bias

Computationally Efficient Optimization of Plackett-Luce Ranking Models for Relevance and Fairness

1 code implementation3 May 2021 Harrie Oosterhuis

Unlike existing approaches that are based on policy gradients, PL-Rank makes use of the specific structure of PL models and ranking metrics.


Robust Generalization and Safe Query-Specialization in Counterfactual Learning to Rank

1 code implementation11 Feb 2021 Harrie Oosterhuis, Maarten de Rijke

We introduce the Generalization and Specialization (GENSPEC) algorithm, a robust feature-based counterfactual LTR method that pursues per-query memorization when it is safe to do so.


Learning from User Interactions with Rankings: A Unification of the Field

no code implementations9 Dec 2020 Harrie Oosterhuis

The goal of a ranking system is to help a user find the items they are looking for with the least amount of effort.


Unifying Online and Counterfactual Learning to Rank

1 code implementation8 Dec 2020 Harrie Oosterhuis, Maarten de Rijke

With the introduction of the intervention-aware estimator, we aim to bridge the online/counterfactual LTR division as it is shown to be highly effective in both online and counterfactual scenarios.

Learning-To-Rank online learning +1

When Inverse Propensity Scoring does not Work: Affine Corrections for Unbiased Learning to Rank

1 code implementation24 Aug 2020 Ali Vardasbi, Harrie Oosterhuis, Maarten de Rijke

Our main contribution is a new estimator based on affine corrections: it both reweights clicks and penalizes items displayed on ranks with high trust bias.


Taking the Counterfactual Online: Efficient and Unbiased Online Evaluation for Ranking

1 code implementation24 Jul 2020 Harrie Oosterhuis, Maarten de Rijke

LogOpt turns the counterfactual approach - which is indifferent to the logging policy - into an online approach, where the algorithm decides what rankings to display.

Selection bias

Policy-Aware Unbiased Learning to Rank for Top-k Rankings

1 code implementation18 May 2020 Harrie Oosterhuis, Maarten de Rijke

We prove that the policy-aware estimator is unbiased if every relevant item has a non-zero probability to appear in the top-k ranking.


FOCUS: Flexible Optimizable Counterfactual Explanations for Tree Ensembles

1 code implementation27 Nov 2019 Ana Lucic, Harrie Oosterhuis, Hinda Haned, Maarten de Rijke

Model interpretability has become an important problem in machine learning (ML) due to the increased effect that algorithmic decisions have on humans.

Unbiased Learning to Rank: Counterfactual and Online Approaches

no code implementations16 Jul 2019 Harrie Oosterhuis, Rolf Jagerman, Maarten de Rijke

Through randomization the effect of different types of bias can be removed from the learning process.


To Model or to Intervene: A Comparison of Counterfactual and Online Learning to Rank from User Interactions

2 code implementations15 Jul 2019 Rolf Jagerman, Harrie Oosterhuis, Maarten de Rijke

At the moment, two methodologies for dealing with bias prevail in the field of LTR: counterfactual methods that learn from historical data and model user behavior to deal with biases; and online methods that perform interventions to deal with bias but use no explicit user models.

Learning-To-Rank online learning +1

Differentiable Unbiased Online Learning to Rank

1 code implementation22 Sep 2018 Harrie Oosterhuis, Maarten de Rijke

Instead, its gradient is based on inferring preferences between document pairs from user clicks and can optimize any differentiable model.

Learning-To-Rank online learning

Ranking for Relevance and Display Preferences in Complex Presentation Layouts

1 code implementation7 May 2018 Harrie Oosterhuis, Maarten de Rijke

Existing learning to rank methods cannot handle such complex ranking settings as they assume that the display order is known beforehand.


Semantic Video Trailers

no code implementations7 Sep 2016 Harrie Oosterhuis, Sujith Ravi, Michael Bendersky

Our approach effectively captures the multimodal semantics of queries and videos using state-of-the-art deep neural networks and creates a summary that is both semantically coherent and visually attractive.

Video Summarization

Cannot find the paper you are looking for? You can Submit a new open access paper.