You are here:
Simulating exposure-related behaviors using agent-based models embedded with needs-based artificial intelligence
Citation:
Brandon, N., K. Dionisio, K. Isaacs, R. Tornero-Velez, D. Kapraun, Woodrow Setzer, AND P. Price. Simulating exposure-related behaviors using agent-based models embedded with needs-based artificial intelligence. Journal of Exposure Science and Environmental Epidemiology . Nature Publishing Group, London, Uk, 30:184–193, (2020). https://doi.org/10.1038/s41370-018-0052-y
Impact/Purpose:
Traditionally, exposure assessors and modelers obtain information about exposure-related behaviors by surveying individuals about their daily activities. Surveyed individuals complete diaries that capture a range of activities over time. However, collecting representative amounts of survey data is difficult and labor-intensive, especially for durations longer than one day. In addition, surveys cannot collect information of which the individual is either unaware (e.g., use of products by another nearby individual or the behavior of children) or is too time-consuming to report (e.g., amounts of each consumer product used in a given day). To avoid these difficulties, we have developed a method that models exposure-related behavior through use of agent-based models. Agent-based models (ABMs) simulate the actions and interactions of autonomous agents within a system. The goal of an ABM is to gain explanatory insights into the behavior of complex processes by simulating the behavior of agents obeying a set of rules. Over time, the interaction between rules and agents can often lead to emergent behavior useful in modeling systems. ABMs combine elements of decision theory, computational sociology, and Monte Carlo methods, and are widely used in a range of research areas including sociology, biology, and ecology . This paper presents a framework that uses methods from both the ABM and the artificial intelligence communities in order to model human behavior for simulating daily activities.
Description:
Exposure to a chemical is a critical consideration in the assessment of risk, as it adds real-world context to toxicological information. Descriptions of where and how individuals spend their time are important for characterizing exposures to chemicals in consumer products and in indoor environments. Herein we create an agent-based model (ABM) that simulates longitudinal patterns in human behavior. By basing the ABM upon an artificial intelligence (AI) system, we create agents that mimic human decisions on performing behaviors relevant for determining exposures to chemicals and other stressors. We implement the ABM in a computer program called the Agent-Based Model of Human Activity Patterns (ABMHAP) that predicts the longitudinal patterns for sleeping, eating, commuting, and working. We then show that ABMHAP is capable of simulating behavior over extended periods of time. We propose that this framework, and models based on it, can generate longitudinal human behavior data for use in exposure assessments.
URLs/Downloads:
DOI: Simulating exposure-related behaviors using agent-based models embedded with needs-based artificial intelligence