Recommender systems and sustainability: a dual perspective

The concept of sustainability, as outlined by the United Nations’ Sustainable Development Goals (SDGs), refers to the ability to meet the needs of the present without compromising the ability of future generations to meet their own needs. This vision is addressed by combining goals concerning the environmental, social, and economic spheres. In this context, Recommender Systems (RS) have emerged as tools that can foster these principles by nudging responsible user behavior and promoting sustainable decision-making. However, the interplay between RS and sustainability is inherently complex since it can be analyzed from two different perspectives: (i) RS for Sustainability, which focuses on how recommendation algorithms can support the achievement of SDGs, and (ii) Sustainability of RS, which focuses on developing recommendation models that inherently adhere to sustainability principles. While the integration of both these perspectives is beneficial and crucial, unfortunately, the current literature has addressed these aspects independently. Accordingly, in this survey, we first provide a comprehensive review of the existing literature on RS that either promotes sustainable behaviors aligned with the SDGs or embeds sustainability principles into their algorithmic design. Next, we identify current gaps and propose key research directions toward an integrated, holistic approach that concurrently addresses both aspects to advance the development of sustainable RS.