Hierarchical Contextual Uplift Bandits for Catalog Personalization

TL;DR

This paper introduces a Hierarchical Contextual Uplift Bandit framework for catalog personalization that adapts to dynamic environments, improving recommendation quality and revenue in fantasy sports platforms.

Contribution

The paper presents a novel hierarchical and uplift modeling approach for contextual bandits, enabling better adaptation and transfer in rapidly changing environments.

Findings

0.4% revenue increase in A/B testing

Improved user satisfaction metrics

Additional 0.5% revenue boost after deployment

Abstract

Contextual Bandit (CB) algorithms are widely adopted for personalized recommendations but often struggle in dynamic environments typical of fantasy sports, where rapid changes in user behavior and dramatic shifts in reward distributions due to external influences necessitate frequent retraining. To address these challenges, we propose a Hierarchical Contextual Uplift Bandit framework. Our framework dynamically adjusts contextual granularity from broad, system-wide insights to detailed, user-specific contexts, using contextual similarity to facilitate effective policy transfer and mitigate cold-start issues. Additionally, we integrate uplift modeling principles into our approach. Results from large-scale A/B testing on the Dream11 fantasy sports platform show that our method significantly enhances recommendation quality, achieving a 0.4% revenue improvement while also improving user satisfaction metrics compared to the current production system. We subsequently deployed this system to production as the default catalog personalization system in May 2025 and observed a further 0.5% revenue improvement.