When Inverse Propensity Scoring does not Work: Affine Corrections for Unbiased Learning to Rank
This addresses a critical bias issue in ranking systems for users, but it is incremental as it builds on existing CLTR frameworks.
The authors tackled the problem of trust bias in Counterfactual Learning to Rank, proving that existing methods, including Inverse Propensity Scoring, fail to correct it, and introduced a new affine correction estimator that removes both trust and position bias, showing in experiments that it approximates the optimal ranking system more closely than before.
Besides position bias, which has been well-studied, trust bias is another type of bias prevalent in user interactions with rankings: users are more likely to click incorrectly w.r.t. their preferences on highly ranked items because they trust the ranking system. While previous work has observed this behavior in users, we prove that existing Counterfactual Learning to Rank (CLTR) methods do not remove this bias, including methods specifically designed to mitigate this type of bias. Moreover, we prove that Inverse Propensity Scoring (IPS) is principally unable to correct for trust bias under non-trivial circumstances. 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. Our estimator is the first estimator that is proven to remove the effect of both trust bias and position bias. Furthermore, we show that our estimator is a generalization of the existing CLTR framework: if no trust bias is present, it reduces to the original IPS estimator. Our semi-synthetic experiments indicate that by removing the effect of trust bias in addition to position bias, CLTR can approximate the optimal ranking system even closer than previously possible.