This article builds on the intersection of educational process mining and the automatic analysis of student's collaborative interaction data previously collected from a web-based multi-tabletop learning environment. The main focus of the article was to analyze and interpret the data using several process mining techniques in order to increase the instructor's awareness (knowledge) about the students' collaboration process and group progress in terms of specific quantitative indicators as follows: participation (consisting of participation density, participation rate and participation dynamics metrics), interaction (consisting of interaction density and interaction dynamics metrics), time performance (including the number of time intervals between the activities as well as the duration of idle/inactive periods), similarity of tasks (or symmetry of actions) and division of labor (or symmetry of roles). The empirical findings showed that there are substantial differences between the high and low performance groups.