Algorithmic Amplification of Politics and Engagement Maximization on Social Media - Archive ouverte HAL Accéder directement au contenu
Pré-Publication, Document De Travail Année : 2023

Algorithmic Amplification of Politics and Engagement Maximization on Social Media

Résumé

This study examines how engagement-maximizing recommender systems influence the visibility of Members of Parliament's tweets in timelines. Leveraging engagement predictive models and Twitter data, we evaluate various recommender systems. Our analysis reveals that prioritizing engagement decreases the ideological diversity of the audiences reached by Members of Parliament and increases the reach disparities between political groups. When evaluating the algorithmic amplification within the general population, engagement-based timelines confer greater advantages to mainstream right-wing parties compared to their left-wing counterparts. However, when considering users' individual political leanings, engagement-based timelines amplify ideologically aligned content. We stress the need for audits accounting for user characteristics when assessing the distortions introduced by personalization algorithms and advocate addressing online platform regulations by evaluating the metrics platforms aim to optimize, beyond the mere algorithmic implementation.
Fichier principal
Vignette du fichier
CNA_23.pdf (1.47 Mo) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-04189507 , version 1 (28-08-2023)

Identifiants

  • HAL Id : hal-04189507 , version 1

Citer

Paul Bouchaud. Algorithmic Amplification of Politics and Engagement Maximization on Social Media. 2023. ⟨hal-04189507⟩
184 Consultations
102 Téléchargements

Partager

Gmail Facebook X LinkedIn More