Mitigating the Position Bias of Transformer Models in Passage Re-Ranking

TL;DR

This paper identifies and addresses position bias in passage re-ranking datasets, proposing a debiasing method that improves the robustness and transferability of Transformer-based models across biased and unbiased datasets.

Contribution

It introduces a novel debiasing technique for passage re-ranking datasets that reduces position bias and enhances model generalization and transfer learning capabilities.

Findings

Debiasing improves model performance on unbiased datasets

Mitigating position bias enhances transfer learning between datasets

Models trained on debiased data perform consistently across biased and unbiased datasets

Abstract

Supervised machine learning models and their evaluation strongly depends on the quality of the underlying dataset. When we search for a relevant piece of information it may appear anywhere in a given passage. However, we observe a bias in the position of the correct answer in the text in two popular Question Answering datasets used for passage re-ranking. The excessive favoring of earlier positions inside passages is an unwanted artefact. This leads to three common Transformer-based re-ranking models to ignore relevant parts in unseen passages. More concerningly, as the evaluation set is taken from the same biased distribution, the models overfitting to that bias overestimate their true effectiveness. In this work we analyze position bias on datasets, the contextualized representations, and their effect on retrieval results. We propose a debiasing method for retrieval datasets. Our results show that a model trained on a position-biased dataset exhibits a significant decrease in re-ranking effectiveness when evaluated on a debiased dataset. We demonstrate that by mitigating the position bias, Transformer-based re-ranking models are equally effective on a biased and debiased dataset, as well as more effective in a transfer-learning setting between two differently biased datasets.