Computational modeling and formal analysis techniques in interdisciplinary studies of complex systems

Computational methods offer new problem solving and analysis techniques that play a key role in advancing the boundaries of many research disciplines. Particularly for social systems, the precision and rigor offered by mathematical models facilitate establishing a clear understanding of the underlying complex system. Computational models also allow for dynamic testing and computer-assisted experiments that may be impossible to carry out in the real world. Modeling something as complex and diverse as a social system is a highly iterative and potentially open-ended process calling for software development techniques that address its specific needs. We present an integrated methodological framework and tool environment for design, validation, and simulation of models of complex social systems. We illustrate consistency and applicability of the framework through novel applications in two different interdisciplinary contexts: Computational Criminology and Identity Management Systems (IMS). The Computational Criminology project, called Mastermind, aims at developing computational models of criminal behavior to facilitate systematic experimental studies of a wide range of criminal activities in urban environments. The Mastermind model, developed in close collaboration with criminologists, focuses on spatial and temporal aspects of different forms of crime. Pushing beyond conventional empirical research, it provides a solid basis for engaging the use of computational thinking and social simulations in crime analysis research and practice. The IMS project, called Identity Management Architecture, aims at consolidating diverse multidisciplinary views on identity management in a systematic fashion. We propose a firm semantic foundation that facilitates a rigorous study of IMS and provides improved accuracy in reasoning about their key properties. The proposed framework is built upon essential concepts of IMS and serves as a starting point for bringing together different approaches in a coherent and consistent manner. We show the value of such a model in the study of semantic aspects of identity theft, unveiling some of the loose ends and ambiguities in the common understanding of the domain. As such, the research presented in this thesis supports the efforts in applying computational modeling techniques to the study of social phenomena by providing a systematic modeling framework and showing its value in novel application contexts