This award will support a workshop that will bring together experts from a broad range of fields to examine and advance the development and use of agent-based models (ABMs) in the social, human-environment, and life sciences. The conference will facilitate community-wide consideration of issues that enhance and broaden the value and utility of agent-based modeling for scholarly and practical purposes. The conference will advance a multi-faceted, community-based examination of current strengths and shortcomings of ABM development and use. The conference and its follow-up products will identify incentives and directions for many related fields to develop more robust, user-friendly tools based on service-oriented ABM designs and platforms. The workshop also has the potential to facilitate interactions between academic-based ABM developers and users and commercial companies in ways that may facilitate advances in ABM software and capabilities. Because of the diversity of fields from which participants are drawn, the conference will promote more interdisciplinary development and use of advanced methodologies.
The use of agent-based models has increased rapidly over the last two decades. The number of articles reporting the development or use of ABMs has increased at an exponential rate since the 1990s, spanning a diverse range of scientific communities that includes geography, ecology, human-environmental science, land-system science, and sociology. A strength of agent-based models has been the potential for representing agent-agent and agent-environment interactions, local-level information and heterogeneity, nonlinearity, feedback, and individual-level activity and decision-making. ABMs also are useful for integrating data and models across multiple systems and for addressing problems across spatial, temporal, and organizational scales. Despite these advances, a number of challenges confront ABM developers and users, including difficulty in model verification and validation, steep learning curves for non-modeling experts, and the lack of common standards or protocols, and limitations regarding their transparency and reusability. To address these challenges, more than three dozen individuals who are ABM modelers and users will gather to identify strengths and weaknesses in the state-of-the-science of agent-based modeling. Participants will engage in a set of discussions on a range of topics that may include model validation, modeling of human decisions, model transparency and reusability, and developing ABMs that can better use "big data." Participants also will identify resources, areas for collaboration, potential tasks, and future directions for the ABM community. Insights and products after the conclusion of the conference are expected to include a synthesis paper and several topical papers dealing with ABMs; a book about ABM science, technology, and application; and an online repository that shares useful ABM resources. Organizers expect this conference to be conducted during the spring months of 2017, most likely in San Diego, California.
