Academic Journals Database
Disseminating quality controlled scientific knowledge

Interventions via Social Influence for Emergent Suboptimal Restraint Use

Author(s): Felicitas MOKOM | Ziad KOBTI

Journal: ADCAIJ : Advances in Distributed Computing and Artificial Intelligence Journal
ISSN 2255-2863

Volume: 1;
Issue: 5;
Start page: 25;
Date: 2013;
Original page

Keywords: Agent-based simulation | Norm emergence | Suboptimal restraint use Social influence

Although restraint use has increased primarily in developed countries, vehicle accident-related injuries and deaths continue to be a problem. Alongside lack of restraint use, studies involving suboptimal restraint use have gained recent popularity. In this study we investigate the use of social influence forinterventions to counter emerging suboptimal restraint use in groups of agents.A multi-agent simulation model is provided where dominant individuals use randomly assigned influence rates to repeatedly alter the knowledge of lessinfluential group members. Cultural influence is implemented via a cultural algorithm and used to simulate individuals affected by beliefs in the community. Objectives include investigating the emergence of patterns of restraint selection and use as well as interventions targeted at more influential agents. Results demonstrate that prominent patterns of behaviour similar to the influentialmembers of the groups do emerge. Furthermore, interventions targeted at influential group members outperform interventions targeted at a percentage of the population at large. Interventions succeed at some level both in the presence and absence of cultural influence.

Tango Jona
Tangokurs Rapperswil-Jona

RPA Switzerland

RPA Switzerland

Robotic process automation