Graph-based analysis of data from human behaviour simulations

In this article, which describes our current work in progress, we demonstrate the potential and universality of graph-based methods for analysing data from computer simulations of human behaviour. The first method, MASDA algorithm, analyses action sequences through an ontology-driven process of abstraction, and aims at isolating strategic sequences and strategic action descriptions in a graphical and symbolic form. The second method, based on the spreading activation algorithm, aims at finding significant correlations between the simulation results and its various internal parameters, settings, events and entities. We illustrate our methods in the context of the EDA project A-0938-RT-GC EUSAS, where they are to be used after validation.