سامانه همایش های دانشگاه آزاد اسلامی واحد نجف آباد

صفحه اصلی / سومین کنفرانس بین المللی مهندسی و علوم کامپیوتر و اولین کنفرانس ملی سامانه های هوشمند، کلان داده و رایانش پیشرفته

Cheating Detection Methods in Online Exams Based on Data Mining and Machine Learning: An Analytical Review

نویسندگان :

Fatemeh Jalali dehkordi¹ Behrang Barekatain²

1- دانشگاه ازاد واحد نجف اباد 2- دانشگاه ازاد نجف اباد

کلمات کلیدی :

cheating detection،online exams،educational data mining،machine learning،anomaly detection

چکیده :

With the expansion of e-learning systems and the widespread use of online exams, cheating detection has become a critical challenge in educational data analysis. The virtual nature of online assessments and the diversity of examinee behaviors have reduced the effectiveness of traditional monitoring methods, increasing the need for data-driven solutions based on educational data mining and machine learning. This paper presents an analytical review of existing studies on cheating detection in online exams, focusing on input data types, feature extraction strategies, and analytical approaches, including statistical analysis, answer similarity measurement, supervised machine learning, and anomaly detection. The analysis shows that machine learning-based methods, particularly those utilizing behavioral and multidimensional data, are more effective in detecting complex cheating patterns, while unsupervised approaches are suitable when labeled data are limited. Despite recent progress, challenges such as model interpretability, false positive risks, and privacy concerns remain. The findings highlight the importance of combined and interpretable data-driven approaches for practical deployment in real-world online exam environments