Friday, November 7, 3:00 p.m.
Center for Social Complexity Suite
Research Hall, Third Floor

Leveraging Crowdsourced data for Computational Social Science: Opportunities, Examples and Challenges

Andrew Crooks, Assistant Professor
Computational Social Science
George Mason University

ABSTRACT: Crowd-sourcing is providing the CSS community with new kinds of data, much of which has a geographic component. While much progress has been made utilizing these new sources of data in parts of the CSS discipline, only recently have they been integrated into agent-based models. This paper will discuss the opportunities that crowd-sourced data provides for agent-based models, specifically how it gives us a new lens to study the micro-interactions of individuals. Through as series of examples, I will demonstrate how such data can be integrated into agent-based models ranging from the distribution of aid after an earthquake, the spread of diseases within refugee camps and slums, to that of evacuations during wild fire events. By building on these examples, I will showcase how the spatial environment and agent populations can be built using crowd-sourced information and how agents’ behaviors can be informed and validated by such information. However, using such data is not without its challenges.