Abstract: Modern society places a high importance on online education because of how quickly technology is developing and how education must change to keep up. E-learning is the only option left following the COVID-19 pandemic to keep instruction going during lockdowns, though. Artificial plays an important role in it. The avoidance of unfair means occurring during online exams is one of the most challenging circumstances exam invigilators encounter. Some of the issues can need consulting nearby references or perhaps getting assistance from neighbours. The principles of facial detection and recognition by Local Binary Pattern Histogram Algorithm, Dlib, toolkit, OpenCV library, and YOLOv3 are used in this research to offer a smart invigilation system that can facilitate exam enrollments and eliminate methods of impersonation and cheating. The evaluation of responses, particularly those of the subjective variety, is one of the main difficulties of online exams. Subjective responses gauge a student's capacity for information retention and verbal expression. Subjective questions, in contrast to objective questions, may have more than one valid response. These responses can state the same thing in a different language and grammatical structure. As a result, grading subjective questions manually takes a lot of time and is difficult to automate. This work uses machine learning (ML) and natural language processing (NLP) to automatically grade subjective questions. The objective response and the ideal response offered by the body that formulated the question were contrasted in the study.

Keywords:  Machine Learning, Natural Language Processing (NLP), artificial intelligence, facial detection

Works Cited:

Rudresh Kale, Aniket Gaikwad, Shivani Gunjal, Prashant Gawali, Abhay R. Gaidhani " Review Online Examination System Using Artificial Intelligence", IJARCCE International Journal of Advanced Research in Computer and Communication Engineering, vol. 12, no. 10, pp. 62-66, 2023. Crossref https://doi.org/10.17148/IJARCCE.2023.121008

PDF | DOI: 10.17148/IJARCCE.2023.121008

Open chat
Chat with IJARCCE