Multi-agent Simulation of Violence Emergence in Protests - Cognition, Perception et Usages
Communication Dans Un Congrès Année : 2024

Multi-agent Simulation of Violence Emergence in Protests

Résumé

This paper presents an agent-based model to understand the emergence of aggressive mob during a protest. We study to what extent (1) protester behavior can lead to violence and (2) protest policing doctrine. The multi-agent architecture relies on the Belief Desire Intention paradigm, enriched with emotions, norms and personality. Indeed, to reproduce phenomena such as tension-building, the protester model includes emotional components that are added to Epstein's model of civil violence. In Epstein's model, aggressive behavior is triggered as a tradeoff between grievance and perceived risk of arrest. Our model is implemented on the open-source GAMA platform and applied to the study of two factors observed in real-life situations: policeprotester ratio and protest configuration (police blockade of escape routes, also called kettling, or police moving in the crowd). Simulations show that the key parameters are not necessarily the police-protester ratio but rather the protest configuration.
Fichier principal
Vignette du fichier
HAR2024.pdf (386.58 Ko) Télécharger le fichier
Origine Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-04842616 , version 1 (17-12-2024)

Licence

Identifiants

  • HAL Id : hal-04842616 , version 1

Citer

Julien Rosenberger, Julien Saunier, Nicolas Sabouret. Multi-agent Simulation of Violence Emergence in Protests. 3rd International Conference on Human and Artificial Rationalities, Sep 2024, Paris, France. ⟨hal-04842616⟩
0 Consultations
0 Téléchargements

Partager

More