Development of An Online Exam Security System using Ensemble Method

For running online exam, it requires a strict security system. Instead using User ID, Password, and token, many security systems used face recognition for authorization. This study proposed an online exam security system that minimized system deficiencies could be used in abusing the implementation of the exam. By adding the filter and implementing algorithms, the system enable distinguish between real faces and fake faces, which are covered by masks or using a photo image. The developed system will stop displaying the questions if there is a fraud. Using multi-face detection, the system can identify the examinee who helped by a 3rd party while doing the exam.