Assisted Model Building in the Social Sciences using Data Driven Simulation
Peter Lee, Ed Ferrari, Catriona Kennedy, Georgios Theodoropoulos, Chris Skelcher
Peter Lee, Ed Ferrari
Centre for Urban and Regional Studies, School of Public Policy, University of Birmingham, UK
Catriona Kennedy, Georgios Theodoropoulos
School of Computer Science, University of Birmingham, UK
Chris Skelcher
Institute for Local Government Studies, School of Public Policy, University of Birmingham, UK
Email address of corresponding author: cmk@cs.bham.ac.uk
The complexity of predicting policy outcomes in social sciences demands sophisticated tools to assist in the decision making process. This paper proposes a framework, which integrates agent-based simulations with data-mining to exploit the advantages of both and overcome their respective limitations. The novelty of the framework is the utilisation of "data driven application simulation" (DDDAS) to enhance the reliability of the agent-based model. A prototype architecture is presented along with its initial application to a housing policy case study.
