Contrastive Learning for Diversity-Aware Product Recommendations in Retail
This addresses the issue of diversity in product recommendations for retail customers, but it is incremental as it builds on existing methods for popularity bias.
The paper tackled the problem of limited item catalog exposure and long-tail distributions in recommender systems for large-scale online retail, introducing a contrastive learning approach that improved catalog coverage while preserving recommendation performance.
Recommender systems often struggle with long-tail distributions and limited item catalog exposure, where a small subset of popular items dominates recommendations. This challenge is especially critical in large-scale online retail settings with extensive and diverse product assortments. This paper introduces an approach to enhance catalog coverage without compromising recommendation quality in the existing digital recommendation pipeline at IKEA Retail. Drawing inspiration from recent advances in negative sampling to address popularity bias, we integrate contrastive learning with carefully selected negative samples. Through offline and online evaluations, we demonstrate that our method improves catalog coverage, ensuring a more diverse set of recommendations yet preserving strong recommendation performance.