Editorial: Reviews in recommender systems: 2022

This editorial introduces the Research Topic “Reviews in recommender systems: 2022,” published in Frontiers in Big Data. The collection brings together nine review articles that tackle both foundational and emerging challenges in recommender systems, including fairness, privacy, multi-stakeholder objectives, and beyond-accuracy metrics such as diversity, serendipity, and trustworthiness.

One of the recurring themes is the shift from pure prediction accuracy to more nuanced system goals. For example, the reviews cover differential privacy in collaborative filtering, fairness not only at the user level but across tourism and recruitment domains, and the complexities that multi-objective systems face when stakeholder interests conflict. It’s not just about what item gets recommended, but about who is impacted and how system choices align with broader ethical, legal, and societal goals.

The editorial also highlights gaps, such as the lack of research on privacy in high-stakes scenarios like job markets or finance, and the scarcity of studies optimizing carousel-based interfaces despite their ubiquity. Several articles argue for more sustainable and trustworthy recommender systems, especially as applications expand to domains tied to the UN Sustainable Development Goals.

Ultimately, this Research Topic signals that recommender system research is maturing. Beyond-accuracy objectives, bias mitigation, and trust factors aren’t just afterthoughts—they’re taking center stage in both academic and practical discussions. For anyone looking to understand where the field is heading—especially in balancing accuracy with fairness, privacy, and stakeholder needs—this editorial and its associated reviews provide a timely roadmap.

@article{kowald2024editorial,
  title = {Editorial: Reviews in recommender systems: 2022},
  author = {Kowald, D. and Yang, D. and Laci\'{c}, E.},
  year = {2024},
  journal = {Frontiers in Big Data},
  volume = {7},
  publisher = {Frontiers Media SA},
  doi = {10.3389/fdata.2024.1384460},
  url = {https://doi.org/10.3389/fdata.2024.1384460},
}