Optimizing Social Network Interventions via Hypergradient-Based Recommender System Design
Marino Kühne, Panagiotis D. Grontas, Giulia De Pasquale, Eindhoven University of Technology, [email protected], Giuseppe Belgioioso, KTH Royal Institute of Technology, [email protected], Florian Dörfler, ETH Zürich, [email protected], John Lygeros, ETH Zürich, [email protected] · 2025
Read Optimizing Social Network Interventions via Hypergradient-Based Recommender System Design by Marino Kühne, Panagiotis D. Grontas, Giulia De Pasquale, Eindhoven University of Technology, [email protected], Giuseppe Belgioioso, KTH Royal Institute of Technology, [email protected], Florian Dörfler, ETH Zürich, [email protected], John Lygeros, ETH Zürich, [email protected] (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-2502-12973.txt.
Description
Marino Kühne and five co-authors tune social networks against polarization in twenty twenty-five. Platforms widen access to ideas yet are blamed for splitting opinions. Handcrafted fixes and heuristics dominate current responses. The authors instead treat the network itself as the design variable.
Interaction weights between users, such as contact frequency, become the controls. User opinions evolve under the classical Friedkin-Johnsen model. The goal is optimizing a metric like polarization reduction. That yields a large, non-convex optimization problem. Their hypergradient-based algorithm is simple, scalable, and accepts varied objectives. Readers in control and network science get a principled alternative: design the recommender's connection strengths directly, rather than patching content rules one by one.
Catalog record
| CREATOR | Marino Kühne, Panagiotis D. Grontas, Giulia De Pasquale, Eindhoven University of Technology, [email protected], Giuseppe Belgioioso, KTH Royal Institute of Technology, [email protected], Florian Dörfler, ETH Zürich, [email protected], John Lygeros, ETH Zürich, [email protected] |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 52 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2502-12973