Extracting User Enabledness from Automated Video Analysis

An objective measure of user performance is important to evaluate an explanation. In the domain of boardgames, this is measureable by analysing a match against an AI opponent. In this thesis, the student will be given a video-dataset of Quarto! matches. The performed moves should be extracted and analysed by student-developed quality guidelines to rank the enabledness and draw conclusions on the quality of the given explanation.