Machine Learning Algorithm to Detect Impersonation in an Essay-Based E-Exam - Version 1.0 - Research

Aug 2, 2026

Upload by Himbo

Essay-based E-exams require multi-paragraph, well-structured responses within e-learning platforms. With reduced supervision during virtual delivery, risks of impersonation and content theft increase, making accurate and cost-effective cheating detection essential. The study develops, trains, and evaluates real-time LSTM, RNN, and GRU algorithms, then benchmarks them against other state-of-the-art models in the same task. Models flag possible impersonation or stolen content when discrepancies exceed a predefined threshold, achieving the highest reported accuracy of 98.6% with the GRU approach.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages13

Reading Time33 min

AccessFree to download

Introduction

Background on E-learning and E-assessment

Need for authentication to reduce academic dishonesty

Proposed Approach (LSTM/RNN/GRU)

Global Journal of Computer Science and Technology: D Neural & Artificial Intelligence Volume 23 Issue 1 Version 1.0 Year 2023 Type: Double Blind Peer Reviewed International Research Journal Publisher: Global Journals Online ISSN: 0975-4172 & Print ISSN: 0975-4350 Machine Learning Algorithm to Detect Impersonation in an EssayBased E-Exam By Mr. Joseph Kombe Samuel, Dr. Peter Ochieng & Dr. Solomon Mwanjele TaitaTaveta University Abstract- Essay-based E-exams require answers to be written out at some length in an E-learning platform. The questions require a response with multiple paragraphs and should be logical and wellstructured. These type of examinations are increasingly becoming popular in academic institutions of higher learning based on the experience of COVID-19 pandemic. Since the exam is mainly done virtually with reduced supervision, the risk of impersonation and stolen content from other sources increases. Due to this, there is need to design cost effective and accurate techniques that are able to detect cheating inan essay based E-exam. In this work we develop, train and evaluate real-time LSTM, RNN and GRU algorithms, and then benchmark the performance of the algorithms against other state-of-the-art models in the same study area of detecting cheating in exam in an E-learning environment. Based on a set threshold, the models alert on possible impersonation or stolen content if the discrepancy exceeds the threshold. The evaluation and benchmarking of the algorithms revealed that our GRU model has the highest accuracy of 98.6% compared to other models in similar studies. Keywords: E-exam, BERT, LSTM, RNN, GRU, Essay, E-learning, machine learning, cheating, education. GJCST-D Classification: FOR Code: 170203 MachineLearningAlgorithmtoDetectImpersonationinanEssayBased EExam Strictly as per the compliance and regulations of: Machine Learning Algorithm to Detect Impersonation in an Essay-Based E-Exam Mr. Joseph Kombe Samuel α , Dr. Peter Ochieng σ & Dr. Solomon Mwanjele ρ Abstract-Essay-based E-exams require answers to be written out at some length in an E-learning platform. The questions require a response with multiple paragraphs and should be logical and well-structured. These type of examinations are increasingly becoming popular in academic institutions of higher learning based on the experience of COVID-19 pandemic. Since the exam is mainly done virtually with reduced supervision, the risk of impersonation and stolen content from other sources increases. Due to this, there is need to design cost effective and accurate techniques that are able to detect cheating in an essay based E-exam. In this work we develop, train and evaluate real-time LSTM, RNN and GRU algorithms, and then benchmark the performance of the algorithms against other state-of-the-art models in the same study area of detecting cheating in exam in an E-learning environment. Based on a set threshold, the models alert on possible impersonation or stolen content if the discrepancy exceeds the threshold. The evaluation and benchmarking of the algorithms revealed that our GRU model has the highest accuracy of 98.6% compared to other models in similar studies. Keywords: E-exam, BERT, LSTM, RNN, GRU, Essay, Elearning, machine learning, cheating, education. I. Introduction E -learning would be a type of learning that takes place through the use of electronic media (Janelli, 2018) . In 1999, it was first used during a seminar on Computer-Based Training (CBT) systems. It's also known as "virtual" or "online" learning. E-learning is becoming a necessary component of modern education, demonstrating the significant role of ICT in the current process of teaching-learning (Soni, 2020) . The growth of online devices has facilitated the delivery of information on E-learning platformsto students, wherever and whenever they need it. (Urosevic, 2019) stated that in the year 2017 there were approximately 23 million new online learn

FAQs

Why is impersonation detection important in essay-based e-exams?

Virtual e-exams often involve reduced supervision, which increases risks of impersonation and stolen responses. Authentication is therefore critical to limit academic dishonesty.

Which machine learning models are developed and evaluated in the study?

The work develops and evaluates real-time LSTM, RNN, and GRU algorithms. These models are trained and assessed for the detection task.

How do the models decide whether impersonation or stolen content occurred?

Each model uses a set threshold to measure discrepancy. When the discrepancy exceeds the threshold, the system alerts on possible impersonation or stolen content.

Aug 2, 2026

Upload by Himbo