Abstract
The widespread integration of Computer-Aided Design (CAD) software into engineering education has transformed instructional practices and assessment strategies, creating new challenges for academic integrity. In parametric modelling environments, plagiarism extends beyond simple geometric duplication to include the reuse of modelling sequences, feature-tree structures, and constraint-based design intentions. Traditional plagiarism-detection tools, primarily designed for text-based documents, are ineffective at capturing the procedural and relational aspects of CAD models. This study introduces a technology-aware plagiarism-detection framework designed explicitly for parametric CAD coursework. The proposed method incorporates multi-level similarity analysis, combining geometric comparisons, feature-tree structure assessments, chronological modelling sequences, and sketch-level constraint evaluations. By examining both the final artefacts and the underlying design intent, the system aims to differentiate between independent problem-solving processes and copied or minimally modified models. The framework was tested on a classroom dataset of student submissions. Results demonstrate enhanced effectiveness in detecting unwarranted similarities while minimising sensitivity to minor variations, thereby supporting fairer and more transparent assessments. Beyond detection, the proposed approach aims to contribute to the development of analysis-based training tools that promote academic integrity in computer-aided design teaching. This study emphasises the importance of understanding plagiarism in CAD environments as a technology-mediated phenomenon, thereby requiring specialised assessment strategies aligned with contemporary engineering pedagogy.

This publication has 19 references indexed in Scilit: