Localization Boosting for Growth Markets: Mitigating Cross-Locale Behavioral Bias in Learning-to-Rank

2026-05-11

AI Summary

This work addresses the degradation in ranking quality for local content outside the United States, a consequence of existing learning-to-rank models being predominantly trained on U.S.-centric user data. To mitigate this issue, the authors propose a multi-objective ranking framework that jointly leverages user click signals, relevance labels generated by vision-language models, and a locality-aware feature augmentation mechanism. This approach effectively disentangles exposure bias from semantic supervision by integrating multiple sources of supervisory signals. Evaluated across five geographic regions, the method significantly enhances the visibility of locally relevant content while preserving semantic relevance, thereby alleviating cross-regional behavioral disparities inherent in conventional ranking systems.

Abstract

Adobe Express is expanding internationally, but the US has a disproportionately large content supply and interaction volume. Learning-to-rank (LTR) models trained primarily on behavioral feedback inherit this imbalance: templates popular in US are over-served in non-US locales. This cross-locale exposure bias suppresses local content discoverability and degrades ranking quality in growth locales. We show that click-only training suppresses semantically informative localization features. Adding vision-language model (VLM) graded relevance labels as auxiliary supervision alongside clicks improves semantic alignment but does not preserve local content visibility. We propose a multi-objective framework combining behavioral supervision, VLM-derived relevance signals, and locale-aware boosting. Across five locales, the resulting model improves relevance while restoring stable localization, demonstrating the importance of disentangling exposure from semantic supervision.

Problem

• cross-locale bias
• learning-to-rank
• localization
• exposure bias
• content discoverability

Innovation

• learning-to-rank
• cross-locale bias
• vision-language model
• multi-objective learning
• localization boosting