Using event based markov model simulation to analyze interactive human behavior

We are often interested in finding underlying patterns while investigating human behavior with interactive systems. Knowledge of these patterns can assist in the design of support tools users of complex information systems. Identification of these patterns in large data sets is, however, difficult and requires powerful analytic techniques. We describe the use of a Markov chain simulation as an exploratory method to analyze interactive human behavior. Using data collected from a longitudinal field study, we demonstrate the usefulness of Markov simulations in exploring the relationship between different human actions, and for reasoning about the mental processes underlying behavior with an interactive system. In this paper we are focused on demonstrating the usefulness and relevance of this methodology for analyzing interactive human behavior, rather than on the specific results from analysis of the data in our case study. In addition to techniques, we describe the relevant software programs researchers and practitioners can use to apply this method in their analysis of relationships and patterns within data.