Traditional evaluation metrics for recommender systems such as accuracy, precision, recall, and user satisfaction primarily focus on system performance from a technical or user-centric perspective. These metrics are essential, however, they often overlook broader implications of recommender systems on society and the environment. Sustainability-aware evaluation metrics address this gap by offering a more holistic assessment framework that takes into account long-term impacts and aligns with the United Nations’ Sustainable Development Goals (SDGs). By incorporating environmental, social, and economic dimensions, sustainability-aware metrics enable researchers and practitioners to evaluate how recommender systems influence areas such as resource consumption, social equity, and economic resilience. This broader perspective is crucial as recommender systems increasingly shape consumer behavior, information exposure, and market dynamics. In this article, we discuss basic sustainability-focused evaluation metrics specifically designed for recommender systems. To evaluate their applicability, we present a case study in which these metrics are applied to a real-world product dataset. With this, we highlight their potential to uncover insights that conventional metrics may miss.
Felfernig et al. (Wed,) studied this question.