Qualitative Evaluation of Spatiotemporal Reasoning in Performances

There are limited capabilities for automatically blocking a scene in a virtual environment today. Current techniques for moving characters include Behavior Markup Language (BML) and motion capture files (mocap). However these methods require either detailed technical knowledge, or are not reusable in more than one scene. In our previous work, we block a play utilizing only a play-script and performed quantitative analyses. In this paper, we apply some of those techniques for blocking a scene (Natural Language Processing), and obtain end-user evaluations of the quality of the spatial positioning of the characters. To accomplish this, we created a new survey to qualitatively evaluate the spatiotemporal reasoning within a performance. With this qualitative tool, we compare our hand-mapped baseline from the 1964 production of Hamlet on Broadway to a natural language processing approach to blocking a performance.