Online Exams Cheating Detection

Online education is important in the COVID-19 pandemic, but online exam at individual homes invites students to cheat in various ways, especially collusion. Researches has proved that 43% of graduate students, 68% of undergraduate students, and 95% of high school students are cheated on assignments or exams. Therefore it was really hard to prevent cheatings in online exams. 🤔

The online exam cheatings can be categorized into 2 areas as cheating on authenticating to exam and cheating while the exam is in progress. When examinee authenticating to the exam their identity should be verified to avoid false entry to exams. But the hardest part is observing cheating during an online exam. The examinee use many tricks such as refer physical notes, google answers, use their mobile phones, asking help from another person to cheat in online exams.

Researches has shown different approaches to detect the online exams cheatings. Let’s have a look into some of the research papers regarding online exam proctoring.👉📝

1. E-exam Cheating Detection System

This research paper published on IJACSA Journal in 2017. It investigated the methods used by students for cheating detection in online exams investigated an online proctoring techniques. E-exam management system is proposed to investigate cheating in online exams using Fingerprint Reader to authenticate the examinee, and Eye Tribe Tracker to continuously guarantee that the examinee is the one who is claiming to be.

Set of sample data collected on average time out of screen and average number of times out of the screen and predicted the label as cheating or non-cheating by plotting and classifying data.

That solution can be implemented over the web for remote exams. The research paper not handling the case of students with special needs. The research paper has stated some future works such as improving the continuous authentication with voice print , face authentication and use of keystrokes as continuous authentication after start an exam.

2. A Visual Analytics Approach to Facilitate the Proctoring of Online Exams

This research paper was published in 2021 at Proceeding of CHI conference. The main aim of this research paper is to present a novel visual analysis approach to proctor online exams by analyzing the exam video records and mouse movement data.

The presented solution detects and visualizes the suspected head (ex: abnormal head rotation, face disappearance from the screen) and mouse movements ( copy and paste, moving the mouse out of the exam web page). The related works were categorized in to three parts : online proctoring methods, mouse movement visualization and head pose analysis. This approach combines human efforts with machine learning techniques for the better accuracy. They have designed a rule based suspected case detection engine to identify suspected cases from video, mouse movements and further estimates the risk of cheating with mathematical calculations. The effectiveness and usefulness of the presented solution was evaluated under 3 different usage scenarios.

This research paper presented some limitations of their solution. Types of cheating cases in their dataset are limited but in real online exams, there are more advanced methods of cheating. For example, using an earphone to listen to answer. Another limitation is the data collection is conducted in the strictest closed-book setting.

As future works they planned to improve their approach for real-time proctoring and further evaluate our system in real world online exams.

3. Automated Online Exam Proctoring

Above research paper was published in 2015 at IEEE research papers. It presented a multimedia analytics system that performs automatic online exam proctoring using webcam, wear cam, and microphone.

The presented system includes six basic components that continuously estimate the key behavior cues: user verification, text detection, voice detection, active window detection, gaze estimation and phone detection. Those features are processed and send to a SVM cheat classifier for getting a cheat decision.

The paper discussed limitations of the solution. It states that they can apply more advanced algorithms for each component, such as the deep learning-based feature representation, typing-based continuous authentication ,face alignment-based pose estimation , upper body alignment, and model personalization, pen detection etc.

4. An Intelligent System For Online Exam Monitoring

This research paper was published in 2016 at IEEE research papers. It presented a comprehensive multi modal system to proctor exams. The inputs from webcam audio, video and active window capture are send to an intelligent rule based inference system to decide if any cheatings happening.

Examinee’s face detected through feature extraction and estimate the head pose to detect cheatings through yaw angle variations, audio presence and active window capture. The proposed system evaluated by comparing malpractice detection using real proctoring vs proposed system.

But it presented a limitation. At the moment, the scope of the project is restricted to a single student scenario. As future improvements it presented that the system can also be implemented using the new video capturing tools such as Bullet Camera, Wireless IP Camera etc.

5. Online Examination System with Cheating Prevention Using Question Bank Randomization and Tab Locking

This research paper was published on 2019 at 4th InCIT. There are limitations and dependency on the quality of internet service for proctoring exams online. Therefore this paper identify the root cause of academic malpractice and presenting an approach to randomize questions and locking the tabs to minimize cheatings. Finally, the validation done by the focus group respondents and end users for effectivity and usability.

The exams are randomized based on fixed weights with the difficulty level of the questions for each examinee is assumed un-changed. That was a limitation of the presented approach. This research presented some future improvements such as including gaze tracking to proctor exams, comparing and visualize the scores of quizzes that will help the proctor to determine what quizzes performed badly.

Online exams have become increasingly popular for course instructors to assess the knowledge of students or other test takers. However, it remains unclear on how to conveniently and effectively proctor online exams. With the literature review of above five papers we can conclude that still the online examination cheatings needs to be improved in the future. ✨✨✨

References

1. E-exam Cheating Detection System (semanticscholar.org)
2. A Visual Analytics Approach to Facilitate the Proctoring of Online Exams | Proceedings of the 2021 CHI Conference on Human Factors in Computing Systems (acm.org)
3. Automated Online Exam Proctoring | IEEE Journals & Magazine | IEEE Xplore
4. An intelligent system for online exam monitoring | IEEE Conference Publication | IEEE Xplore
5. Online Examination System with Cheating Prevention Using Question Bank Randomization and Tab Locking | IEEE Conference Publication | IEEE Xplore