Estimating Position Bias without Intrusive Interventions

Abstract

Presentation bias is one of the key challenges when learning from implicit feedback in search engines, as it confounds the relevance signal. While it was recently shown how counterfactual learning-to-rank (LTR) approaches \cite{Joachims/etal/17a} can provably overcome presentation bias when observation propensities are known, it remains to show how to effectively estimate these propensities. In this paper, we propose the first method for producing consistent propensity estimates without manual relevance judgments, disruptive interventions, or restrictive relevance modeling assumptions. First, we show how to harvest a specific type of intervention data from historic feedback logs of multiple different ranking functions, and show that this data is sufficient for consistent propensity estimation in the position-based model. Second, we propose a new extremum estimator that makes effective use of this data. In an empirical evaluation, we find that the new estimator provides superior propensity estimates in two real-world systems -- Arxiv Full-text Search and Google Drive Search. Beyond these two points, we find that the method is robust to a wide range of settings in simulation studies.

Cited by in corpus (15)

• Controlling Fairness and Bias in Dynamic Learning-to-Rank
• CauseRec: Counterfactual User Sequence Synthesis for Sequential Recommendation
• Unifying Online and Counterfactual Learning to Rank
• When Inverse Propensity Scoring does not Work: Affine Corrections for Unbiased Learning to Rank
• Towards Fair Recommendation in Two-Sided Platforms
• Maximizing Marginal Fairness for Dynamic Learning to Rank
• Deep Position-wise Interaction Network for CTR Prediction
• Fairness of Exposure in Light of Incomplete Exposure Estimation
• Taking the Counterfactual Online: Efficient and Unbiased Online Evaluation for Ranking
• U-rank: Utility-oriented Learning to Rank with Implicit Feedback
• Accelerated Convergence for Counterfactual Learning to Rank
• Implicit Feedback for Dense Passage Retrieval: A Counterfactual Approach
• De-Biased Modelling of Search Click Behavior with Reinforcement Learning
• Robust Generalization and Safe Query-Specialization in Counterfactual Learning to Rank
• A General Framework for Pairwise Unbiased Learning to Rank