Calibrated Recommendations: Survey and Future Directions Journal: TORS Volume: 1 1 1 1 CCS: Information systems Recommender systems CCS: Computing methodologies Machine learning
Diego Corrêa da Silva, Dietmar Jannach · 2025
Read Calibrated Recommendations: Survey and Future Directions Journal: TORS Volume: 1 1 1 1 CCS: Information systems Recommender systems CCS: Computing methodologies Machine learning by Diego Corrêa da Silva, Dietmar Jannach (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2507-02643.txt.
Description
Diego Corrêa da Silva and Dietmar Jannach survey a decade of work on calibrated recommendations, in a 2025 paper. A recommender is calibrated when the mix of items it suggests matches the mix of tastes in a user's own history. Without calibration, systems drift toward a narrow slice of what a person likes. Researchers have lately used calibration to fight sameness, bias, and unfairness in suggestions.
The survey reviews the technical recipes for achieving calibration, from post-processing suggestion lists to training with calibration in mind. It also gathers the evidence on whether calibration helps in practice, and for which goals. A closing section lists the limits and snags that builders meet when shipping calibration in real products. The paper serves as a map for a fast-growing corner of recommender research.
Catalog record
| CREATOR | Diego Corrêa da Silva, Dietmar Jannach |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 84 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2507-02643