A Framework for Generating Geospatial Social Computing Environments

Computational social science plays an important role in emergency management from a quantitative perspective. Reconstructing an individual-based social computing environment is crucial for both accurate computational experiments and determining optimal decisions. Here, the authors propose a formalization method to define basic componential models in the artificial society and the inner logic in these models. A detailed generation process is presented, and multisource statistical data, social interactive behavior, and multilayer social networks are integrated together. As an evaluation of the proposed framework, a virtual city of Beijing is reconstructed. Each citizen is endowed with demographic attributes, including age, gender, social role, correlated geographic locations, and multiple social relations. The generated synthetic population is statistically equivalent to the real population. Transmission experiments of influenza are performed in the reconstructed computational environment, and individual daily interacting behavior is tracked and analyzed. The results indicate that the framework can provide an effective methodology to reconstruct the computing environment in high resolution by using low-resolution statistical data, leading to better prediction and management of emergencies.