HEVAH Project

HEVA is a 3D animated character developed with Blender that evaluates the knowledge of a user in a particular subject using an adaptive test. The avatar was built with the Game Engine of Blender 2.25 and Python. Our adaptive test is based on XML as data manager and CLIPS as inference engine. Both are connected with a new library in Python that we had to develop for this purpose. Some characteristic of our project include: • Mesh deformation in real time (dynamic creation and edition of IpoCurves in Game Engine). • Mesh deformation in real time (dynamic creation and edition of IpoCurves in Game Engine). • Interaction with external applications to Blender (such as CLIPS) using pipes. • Use of threads for concurrent execution of tasks. • Use of Text to Speech. Our system is formed by three fundamental modules: • Question Manager (QM): responsible of select and evaluate questions for the adaptive test. It has been developed in Python. • User Interface (UI): compose by the avatar developed in Blender. • Inference Engine (IE): responsible of select the level of knowledge of user through the process of rules in Clips.