Agents – the ‘atomic unit’ of social systems? @AAG 2017

Posted on Monday 27 February 2017 by Andrew Crooks

As part of the Symposium on Human Dynamics in Smart and Connected Communities at the forthcoming AAG Annual Meeting in Boston we have organized 2 sessions under the title of “Agents – the ‘atomic unit’ of social systems?” (session IDs 4169 & 4269). These will be held on on Saturday, 4/8/2017, from 8:00 am to 11.40 (we did not chose this time slot). Below you can see the session description and the list of speakers and titles. We hope some of the readers of this blog can make it to the sessions.

Session Description

By defining a social system as a collection of agents, individuals and their behaviors/decisions become the driving force of these systems. Complex global phenomena such as collective behaviors, extensive spatial patterns, and hierarchies are manifested through agent interaction in such a way that the actions of the parts do not simply sum to the activity of the whole. This allows unique perspectives into the inner workings of social systems, making agent-based modelling (ABM) a powerful and appealing tool for understanding the drivers of these systems and how they may change in the future.

What is noticeable from recent applications of ABM is the increase in complexity (richness and detail) of the agents, a factor made possible through new data sources and increased computational power. While there has always been ‘resistance’ to the notion that social scientists should search for some ‘atomic element or unit’ of representation that characterizes the geography of a place, the shift from aggregate to individual mark agents as a clear contender to fulfill the role of ‘atom’ in social simulation modelling. However, there are a number of methodological challenges that need to be addressed if ABM is to fully realize its potential and be recognized as a powerful tool for policy modelling in key societal issues. Most pressing are methods to accurately identify, represent, and evaluate key behaviors and their drivers in ABM.

This session will present papers that contribute towards this wide discussion ranging from epistemological perspectives of the place of ABM, extracting behavior from novel and established data sets to new, intriguing applications to establishing robustness in calibrating and validating ABMs.

Organizers:

Andrew Crooks, Department of Computational and Data Sciences, George Mason University.
Alison Heppenstall, School of Geography, University of Leeds.
Nick Malleson, School of Geography, University of Leeds
Paul Torrens, Department of Computer Science and Engineering, Tandon School of Engineering, New York University.
Sarah Wise, Centre for Advanced Spatial Analysis (CASA), University College London.