MIchael’s Research

Home My Research My Classes Bio and CV Contact Me

At NC State University, my team and I work to develop computational models of interactive narrative with applications to computer games, sense-making support, educational and training systems and virtual environments.  My work is grounded in computational approaches, but draws heavily on disciplines where human cognition and interaction are central.

A science of narrative in the context of games

Herbert Simon’s Sciences of the Artificial [1] sets the conceptual groundwork for scientific approaches to artificial phenomena.  Like natural sciences, sciences of the artificial seek to define invariant properties between constituent entities of analysis.  Much of my work seeks to contribute to a science of narrative. Narrative is a phenomenon involved in aspects of human activity as diverse and as central as sense-making, communication and play and increasingly plays a role in commercial, cultural and artistic contexts within digital media and virtual worlds.  In my framing, a science of narrative seeks to characterize invariant relationships between inner environment (a reader’s cognitive states), interface (narrative discourse) and outer environment (stories) in order to understand how the design of narratives prompts what we know as narrative experiences.  My approach to this effort is to develop novel computational approaches to modeling narrative phenomena, most typically in the context of the creation of new types of user experiences in computational media like video games.

In my work, I draw on artificial intelligence (AI) models of reasoning about action and change, natural language processing and intelligent user interfaces.   While my work is computational, it’s positioned at the boundary between a virtual world and a human’s understanding of that world.  Reflecting this liminal position, my work has strong humanist elements, making connections to disciplines as divergent from computer science as narrative theory, cognitive psychology and linguistics.  Much of my work involves collaborations with colleagues from these disciplines, engaging in efforts to develop shared vocabularies and common models.

One methodological element that’s central to all work that I do is that of empirical evaluation.  Motivated by Hanks, Pollack and Cohen’s [2] articulation of the role of experimentation in the development of new theories of artificial intelligence, my research follows an iterative process involving theory formation, system construction and controlled experimentation.  Based on the interpretation of my experimental results, I form new theories or revise old ones.  This approach most often leads to the construction of game-like playable artifacts designed to serve both as experimental materials and demonstrations of the capabilities/boundaries of the theory put to practice.  With my students, I have built games that automatically create stories with compelling conflict, games that adapt their stories to unexpected user activity, systems that produce textual narratives that account for coherent intentionality of their characters and systems that automatically select and compose shot sequences within a 3D virtual film set.  We have built and evaluated large-scale serious games used for crime scene investigation, intelligence analysis and STEM-based middle-school education.

Beyond my own research activity, I’ve worked at NC State and in the larger research community to build support for the same type of interdisciplinary research around computational media.  I was a co-founder of two of the strongest venues for research publications around computational games research (the AAAI Conference on Artificial Intelligence and Interactive Digital Entertainment and the Foundations of Digital Games Conference) and served for two years as the editor-in-chief of one of the earliest peer-reviewed journals on games research (the Journal of Game Development).  At NC State, I founded the Digital Games Research Center in 2008, which has now grown to be the focal point for NC State scholars’ work in research and education in games, engaging 19 faculty and staff and over 30 graduate and undergraduate students across four colleges and six departments.  In 2014-2015, the center managed $4.8M in new and continuing awards.  Sponsorship for our research comes from a diverse range of sources that includes the Department of Defense, the National Science Foundation and the Institute for Museum and Library Services.


  1. Simon, H.,  Sciences of the Artificial, MIT Press, third edition, 1996.
  2. Hanks, S., Pollack,. M., and Cohen, P., Benchmarks, test beds, controlled experimentation, and the design of agent architectures, AI Magazine, Volume 14 Issue 4, Winter 1993, Pages 17-42, American Association for Artificial Intelligence Menlo Park, CA, USA