Trading Engagement for Sustainability: Carbon-Aware Re-ranking for E-commerce Recommendations

Unofficial AI-generated summary based on the public title and abstract. Not an official translation.

🔬研究者:This paper provides a novel method for carbon footprint estimation at scale using LLMs and demonstrates the engagement-carbon tradeoff in recommendation systems, offering a benchmark for future research in sustainable AI and e-commerce.

🏢実務担当者:E-commerce platforms can adopt the carbon-aware re-ranking to reduce product-level emissions with minimal impact on engagement, using the LLM-based PCF estimation pipeline to overcome data scarcity.

🏛政策担当者:The findings support policies promoting low-carbon consumption and demonstrate the feasibility of integrating product carbon footprints into digital platforms, relevant for climate disclosure and sustainable finance frameworks.

E-commerce recommender systems strongly influence which products users consider and purchase, yet sustainability signals such as Product Carbon Footprint (PCF) are almost never available at catalog scale. We study carbon-aware product recommendation in the realistic setting where PCF labels are missing for most items and must be inferred. We first estimate product-level carbon footprints via a retrieval-augmented PCF estimation pipeline that transfers supervision from the Carbon Catalogue, a small set of life-cycle-assessed products, to a large unlabeled e-commerce catalog using semantic similarity search, few-shot LLM prompting, and a nearest-neighbour fallback. We then apply a carbon-aware post-hoc re-ranking strategy on top of relevance scores produced by three established recommendation models: BPR, NeuMF, and LightGCN. The method trades off predicted user-item engagement against estimated carbon footprint through a single tunable parameter, lambda. In this offline study, engagement is operationalized through Amazon review interactions, which serve as implicit feedback and as a proxy for user interest or purchase behavior. We evaluate the framework on the Amazon Reviews dataset across three product categories: Home and Kitchen, Sports and Outdoors, and Electronics. By sweeping lambda, we construct Pareto frontiers that characterize the achievable engagement and carbon trade-off for each model and category. Substantial carbon reductions are achievable at minimal engagement cost across all models and categories. However, the available carbon headroom varies by model and category, underscoring the importance of model choice and domain context.

• openalex https://arxiv.org/abs/2606.04550first seen 2026-06-07 04:53:38