A Case Study on the Use of Model Quality Testing Tools for the Assessment of MCAD Models and Drawings

Abstract

[EN] In this paper, we build on the idea that specialized instruction improves the overall quality of CAD documents by guiding students into selecting the most suitable modeling strategies and approaches. To this end, automatic assessment tools can be used to detect errors and provide feedback, thus relieving instructors from routine checks and allowing them to address quality errors and modeling aspects of higher semantic level. A representative commercial Model Quality Testing (MQT) solution was selected as a case study to determine whether these tools may become automated assistants for student evaluation and feedback. As a result, a new taxonomy of modeling aspects that can be automatically checked is proposed. We claim that current MQT tools can supplement the learning of quality concepts, but require significant tuning and only provide limited testing and tutoring capabilities. Extending the capabilities of these tools (through macros or dedicated API¿s), or even developing entirely newMQT tools specifically aimed at instruction purposes, is an essential requirement to develop automated teaching-assistants based on MQT techniques.