An agent-based model of alcohol use and abuse: SimARC

Abstract—Alcohol-related problems (assaults, accidents and/or

crimes) and alcohol abuse are recurrent societal problems

leading to high social costs. Finding adapted policies to tackle

this issue isn’t a trivial task due to the highly complex nature of

alcohol consumption as many interrelated risk factors interact

in a hardly predictable way. This paper describes an agentbased

simulation model, called SimARC (Simulation of

Alcohol-Related Consequences), aiming at exploring the

complex interplay of these factors following a generative

process whereby theory and model co-evolve within iterative

loops. To explore the complexity of alcohol use and abuse, we

need not only to include the aforementioned risk factors but

also their evolution and highly dynamical interactions across

scales. Therefore, our agent-based model aims to encapsulate

several levels of reality. Considering an ontology as catalog of

elements and relation amongst those elements, our ontologydriven

behavioral model includes: neuro-biological responses

to alcohol use (individual level), peer influence channeled

through various social networks (meso-level) and societal

responses to alcohol-related problems (meta-level). This

ontological framework aims to establish a robust test-bed to

analyze – in silico – the plausible consequences of various

public policies related to alcohol abuse in public venues. After

a brief review of the literature, we present SimARC’s core

structure and preliminary results.

Keywords-component; agent-based model, ontology, alcohol,

social simulation, public health