Data driven simulation to support model building in the social sciences

Abstract

Artificial intelligence (AI) can contribute to the management of a data driven simulation system, in particular with regard to adaptive selection of data and refinement of the model on which the simulation is based. We consider two different classes of intelligent agent that can control a data driven simulation: (a) an autonomous agent using internal simulation to test and refine a model of its environment and (b) an assistant agent managing a data-driven simulation to help humans understand a complex system (assisted model-building). We present a prototype implementation of an assistant agent to apply DDDAS to social simulations. The automation of the data-driven model development requires content interpretation of both the simulation and the corresponding real-world data. The paper discusses the use of Association Rule Mining to produce general logical statements about simulation and data content as well as the use of logical consistency checking to detect observations that refute the simulation predictions. Finally we consider ways in which this kind of assistant agent can cooperate with autonomous data collection and analysis agents to build a more complete and reliable picture of the observed system.