DJATI, ATILLA KUNCORO (2026) IMPLEMENTASI DAN EVALUASI SISTEM DETEKSI KECURANGAN KUIS ONLINE BERBASIS PEMANTAUAN AKTIVITAS BROWSER DAN KEYSTROKE DYNAMICS. S1 thesis, Universitas Mercu Buana Jakarta.
|
Text (HAL COVER)
Cover.pdf Download (361kB) | Preview |
|
![[img]](https://repository.mercubuana.ac.id/style/images/fileicons/text.png) |
Text (BAB I)
BAB 1.pdf Restricted to Registered users only Download (116kB) |
|
|
Text (BAB II)
BAB 2.pdf Restricted to Registered users only Download (246kB) |
|
|
Text (BAB III)
BAB 3.pdf Restricted to Registered users only Download (252kB) |
|
|
Text (BAB IV)
BAB 4.pdf Restricted to Registered users only Download (346kB) |
|
|
Text (BAB V)
BAB 5.pdf Restricted to Registered users only Download (90kB) |
|
|
Text (DAFTAR PUSTAKA)
Daftar Pustaka.pdf Restricted to Registered users only Download (95kB) |
|
|
Text (LAMPIRAN)
Lampiran.pdf Restricted to Registered users only Download (503kB) |
Abstract
Online quizzes within Learning Management Systems (LMS) are vulnerable to various forms of cheating, such as switching tabs, pasting text from external sources, or having another person take the quiz. Conventional anti-cheating mechanisms rely on a single class of signals, making them easy to circumvent and less fair to honest participants. This research implements and evaluates a cheating�indication detection system for online quizzes that fuses browser activity monitoring with keystroke dynamics analysis on the EduSkill LMS built on Laravel. Browser activity is monitored through tab switching, window blur, fullscreen exit, and copy-paste attempts, while typing patterns are analyzed using z-score anomaly scoring against each participant's baseline. Both signals are fused into a single integrity risk score ranging from 0 to 100 with low, medium, and high categories. Evaluation was conducted through honest and cheating participant scenarios to measure accuracy, precision, recall, and false positive rate. The results show that fusing both signals improves detection sensitivity without burdening honest participants, while clarifying the system's limitations under cold-start conditions and potential baseline poisoning for future development. Keywords: cheating detection, online quiz, browser activity, keystroke dynamics, learning management system. Online quizzes within Learning Management Systems (LMS) are vulnerable to various forms of cheating, such as switching tabs, pasting text from external sources, or having another person take the quiz. Conventional anti-cheating mechanisms rely on a single class of signals, making them easy to circumvent and less fair to honest participants. This research implements and evaluates a cheating�indication detection system for online quizzes that fuses browser activity monitoring with keystroke dynamics analysis on the EduSkill LMS built on Laravel. Browser activity is monitored through tab switching, window blur, fullscreen exit, and copy-paste attempts, while typing patterns are analyzed using z-score anomaly scoring against each participant's baseline. Both signals are fused into a single integrity risk score ranging from 0 to 100 with low, medium, and high categories. Evaluation was conducted through honest and cheating participant scenarios to measure accuracy, precision, recall, and false positive rate. The results show that fusing both signals improves detection sensitivity without burdening honest participants, while clarifying the system's limitations under cold-start conditions and potential baseline poisoning for future development. Keywords: cheating detection, online quiz, browser activity, keystroke dynamics, learning management system. Kuis daring pada Learning Management System (LMS) rentan terhadap berbagai bentuk kecurangan, seperti berpindah tab, menempel teks dari sumber luar, hingga pengerjaan oleh pihak lain. Mekanisme anti-kecurangan yang umum masih bertumpu pada satu jenis sinyal sehingga mudah dielakkan dan kurang adil bagi peserta jujur. Penelitian ini mengimplementasikan dan mengevaluasi sistem deteksi indikasi kecurangan kuis daring yang memadukan pemantauan aktivitas browser dan analisis keystroke dynamics pada LMS EduSkill berbasis Laravel. Aktivitas browser dipantau melalui perpindahan tab, kehilangan fokus jendela, keluar layar penuh, serta upaya salin-tempel, sedangkan pola mengetik dianalisis menggunakan skor z terhadap baseline peserta. Kedua sinyal difusikan menjadi satu skor risiko integritas berskala 0 sampai 100 dengan kategori rendah, sedang, dan tinggi. Evaluasi dilakukan melalui skenario peserta jujur dan curang untuk mengukur akurasi, precision, recall, dan false positive rate. Hasil pengujian menunjukkan bahwa fusi kedua sinyal meningkatkan kepekaan deteksi tanpa membebani peserta jujur, sekaligus memperjelas keterbatasan sistem pada kondisi cold-start dan potensi baseline poisoning sebagai perhatian pengembangan lanjutan. Kata kunci: deteksi kecurangan, kuis online, aktivitas browser, keystroke dynamics, learning management system.
Actions (login required)
 |
View Item |
