Optimizing Social Network Interventions via Hypergradient-Based Recommender System Design

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Kühne, Marino, Grontas, Panagiotis D., De Pasquale, Giulia, Belgioioso, Giuseppe, Dörfler, Florian, Lygeros, John
Format:	Preprint
Published:	2025
Subjects:	Social and Information Networks Systems and Control Optimization and Control
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

_version_	1866929719069376512
author	Kühne, Marino Grontas, Panagiotis D. De Pasquale, Giulia Belgioioso, Giuseppe Dörfler, Florian Lygeros, John
author_facet	Kühne, Marino Grontas, Panagiotis D. De Pasquale, Giulia Belgioioso, Giuseppe Dörfler, Florian Lygeros, John
contents	Although social networks have expanded the range of ideas and information accessible to users, they are also criticized for amplifying the polarization of user opinions. Given the inherent complexity of these phenomena, existing approaches to counteract these effects typically rely on handcrafted algorithms and heuristics. We propose an elegant solution: we act on the network weights that model user interactions on social networks (e.g., frequency of communication), to optimize a performance metric (e.g., polarization reduction), while users' opinions follow the classical Friedkin-Johnsen model. Our formulation gives rise to a challenging large-scale optimization problem with non-convex constraints, for which we develop a gradient-based algorithm. Our scheme is simple, scalable, and versatile, as it can readily integrate different, potentially non-convex, objectives. We demonstrate its merit by: (i) rapidly solving complex social network intervention problems with 3 million variables based on the Reddit and DBLP datasets; (ii) significantly outperforming competing approaches in terms of both computation time and disagreement reduction.
format	Preprint
id	arxiv_https___arxiv_org_abs_2502_12973
institution	arXiv
publishDate	2025
record_format	arxiv
spellingShingle	Optimizing Social Network Interventions via Hypergradient-Based Recommender System Design Kühne, Marino Grontas, Panagiotis D. De Pasquale, Giulia Belgioioso, Giuseppe Dörfler, Florian Lygeros, John Social and Information Networks Systems and Control Optimization and Control Although social networks have expanded the range of ideas and information accessible to users, they are also criticized for amplifying the polarization of user opinions. Given the inherent complexity of these phenomena, existing approaches to counteract these effects typically rely on handcrafted algorithms and heuristics. We propose an elegant solution: we act on the network weights that model user interactions on social networks (e.g., frequency of communication), to optimize a performance metric (e.g., polarization reduction), while users' opinions follow the classical Friedkin-Johnsen model. Our formulation gives rise to a challenging large-scale optimization problem with non-convex constraints, for which we develop a gradient-based algorithm. Our scheme is simple, scalable, and versatile, as it can readily integrate different, potentially non-convex, objectives. We demonstrate its merit by: (i) rapidly solving complex social network intervention problems with 3 million variables based on the Reddit and DBLP datasets; (ii) significantly outperforming competing approaches in terms of both computation time and disagreement reduction.
title	Optimizing Social Network Interventions via Hypergradient-Based Recommender System Design
topic	Social and Information Networks Systems and Control Optimization and Control
url	https://arxiv.org/abs/2502.12973