Green Recommender Systems: A Call for Attention
Joeran Beel, Alan Said, Tobias Vente, Lukas Wegmeth · ACM SIGIR Forum · 2024
The computational demands of recommender systems have drastically increased over the last decade, leading to higher energy consumption and carbon emissions. As recommender systems become more and more central to industries worldwide, their environmental impact is growing rapidly. However, this challenge also presents an opportunity. Designing recommender systems to minimize energy consumption and reduce environmental costs - this is what we define as Green Recommender Systems - can offer direct benefits not only for sustainability but also for companies and users. By optimizing resource use, these systems can reduce operational costs, improve efficiency, and even enhance user experience through faster recommendations. This paper calls for the recommender systems community to urgently integrate energy efficiency into their designs, considering it alongside traditional performance metrics. Addressing the environmental impact of recommender systems is essential, not just for the planet, but also for the long-term viability and cost-effectiveness of the technology.