Alibaba Cloud Developer

Dec 12, 2018 · Artificial Intelligence

Tackling Pseudo-Exposure in Mobile E-Commerce: A Contextual Multiple-Play Bandit Approach

To address the pseudo-exposure problem that reduces click-through rates in mobile e-commerce recommendation, the authors model the task as a contextual multiple-play bandit, propose weighted sample and similarity-enhanced linear reward extensions, provide sublinear regret proofs, and demonstrate significant CTR gains on real Taobao data.