Correcting Exposure Bias for Link Recommendation
This addresses bias and generalization issues in recommender systems for domains like academic citations, though it is incremental as it builds on existing bias correction techniques.
The paper tackles exposure bias in link recommendation systems, such as in citation networks, where systematic underexposure to relevant items leads to biased predictions and feedback loops; experiments on semi-synthetic data show that their methods reliably identify truly relevant citations and increase diversity in recommended papers' fields.
Link prediction methods are frequently applied in recommender systems, e.g., to suggest citations for academic papers or friends in social networks. However, exposure bias can arise when users are systematically underexposed to certain relevant items. For example, in citation networks, authors might be more likely to encounter papers from their own field and thus cite them preferentially. This bias can propagate through naively trained link predictors, leading to both biased evaluation and high generalization error (as assessed by true relevance). Moreover, this bias can be exacerbated by feedback loops. We propose estimators that leverage known exposure probabilities to mitigate this bias and consequent feedback loops. Next, we provide a loss function for learning the exposure probabilities from data. Finally, experiments on semi-synthetic data based on real-world citation networks, show that our methods reliably identify (truly) relevant citations. Additionally, our methods lead to greater diversity in the recommended papers' fields of study. The code is available at https://github.com/shantanu95/exposure-bias-link-rec.