Article Per Year (5 Year)
p-Index From 2021 - 2026
0.562
P-Index
This Author published in this journals
All Journal Jurnal Nasional Teknologi Komputer
Harmiati Bungsu Bangun
Universitas Pembangunan Panca Budi
Computer Science & IT

Published : 3 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 3 Documents
Search

 1 
PENERAPAN ALGORITMA RANDOM FOREST DALAM MENGKLASIFIKASIKAN KELULUSAN TEPAT WAKTU MAHASISWA PASCASARJANA TEKNOLOGI INFORMASI STAMBUK 2023 UNIVERSITAS PEMBANGUNAN PANCA BUDI Ade Surya Bakti Pane; Harmiati Bungsu Bangun; Astri Mutia Rahma; Rian Farta Wijaya
Jurnal Nasional Teknologi Komputer Vol 6 No 3 (2026): Juli 2026
Publisher : CV. Hawari

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.61306/jnastek.v6i3.477

Abstract

Timely graduation is one of the indicators of success in higher education administration. Students in the Master of Information Technology Program are required to complete all academic obligations, including fulfilling study-credit requirements and completing a thesis, within the normal study period of two years or four semesters. This study aims to apply the Random Forest algorithm to classify timely graduation among Master of Information Technology students from the 2023 cohort at Universitas Pembangunan Panca Budi, analyze the model performance, and identify the academic attributes that contribute most to the classification results. The study used academic data from 27 students, including total completed credits, Cumulative Grade Point Average (CGPA), and graduation status. Graduation status was categorized into timely graduation and delayed graduation. The data were processed through preprocessing, label encoding, an 80:20 training-testing data split using stratified split, and Random Forest model development with 100 decision trees. The results showed that the model achieved an accuracy of 83.33%, precision of 100%, recall of 80.00%, and an F1-score of 88.89%. The feature importance analysis indicated that CGPA contributed 100% to the classification results, while completed credits contributed 0% because all students completed the same number of credits, namely 48 credits. Therefore, the Random Forest algorithm can be used to classify students’ timely graduation based on the available academic data.
Analisis Deteksi Kecurangan Ujian Online Dengan Sistem Pakar Berbasis Event Browser Tanpa Kamera : Studi Kasus Penyelenggara Olimpiade Sains Digniti Ade Surya Bakti Pane; Harmiati Bungsu Bangun; Astri Mutia Rahma; Erbin Sitorus; Rian Farta Wijaya; Zulham Sitorus
Jurnal Nasional Teknologi Komputer Vol 6 No 1 (2026): Januari 2026
Publisher : CV. Hawari

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.61306/jnastek.v6i1.337

Abstract

The implementation of online exams differs significantly from exams supervised directly by proctors. Participants' body movements and interactions can be easily read. Even with camera monitoring, participants can still easily switch browser tabs or open applications on their devices. On the other hand, camera-based proctoring solutions can also raise invasive privacy concerns. Furthermore, device limitations can hinder someone from taking the exam due to the lack of a camera, and bandwidth limitations can result in unstable connections during the exam. Sample log activity data from test participants for log_id ranges from 44790 to 46883, while the actions consist of fullscreen_exit, context_switch, standby_log, focus_return, fullscreen_enter, and devtools_open. The results and rules of this test have scores, with each result having an average value of 1, and the rules having an average value of 1-10. In this process, experts assess each detection result using a binary scheme: a score of 1 is given if the expert system's conclusion matches the expert's decision, and a score of 0 is given if the expert system's decision does not match the expert's decision. This study proposes an expert system framework with a cameraless forward chaining method running on the Google Chrome browser (desktop and Android) as decision support. The system utilizes browser event telemetry (visibility changes, fullscreen, standby) that aligns with existing policies, such as mandatory fullscreen mode, copy/paste blocking, and question blurring when exiting fullscreen or a visibility change occurs. This can generate an audit trail that can be audited by decision makers.
Pengembangan Protokol TEBAS (Teliti, Evaluasi, Blokir, Amankan, Sampaikan) sebagai Kerangka Mitigasi Serangan Social Engineering Berbasis Komunitas Khairul; Ade Surya Bakti Pane; Harmiati Bungsu Bangun; Astri Mutia Rahma; Erbin Sitorus
Jurnal Nasional Teknologi Komputer Vol 6 No 1 (2026): Januari 2026
Publisher : CV. Hawari

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.61306/jnastek.v6i1.343

Abstract

In the contemporary cybersecurity landscape, the attack paradigm shift from exploiting technical vulnerabilities to manipulating human psychology (social engineering) has reached a critical level. The 2024 and 2025 annual reports from various global and national security authorities indicate that traditional defense mechanisms centered on the technical perimeter are no longer sufficient to stem the tide of attacks targeting the user's cognitive layer. This paper proposes the development and formalization of a new mitigation protocol, TEBAS (Research, Evaluation, Block, Secure, Deliver). This protocol was designed using the Design Science Research (DSR) methodology to address the specific needs of Indonesia's digital ecosystem, characterized by a highly communal culture but facing a significant digital literacy gap. TEBAS integrates human-centric cybersecurity principles with global standard frameworks such as the NIST Cybersecurity Framework 2.0, but is adapted into operational procedures that can be implemented by individuals and communities. Through an in-depth analysis of current attack vectors such as Quishing (QR Code Phishing) and Deepfake Vishing, and validation using a modified Delphi method, this study demonstrates that TEBAS offers a holistic approach that focuses not only on individual detection but also on collective resilience through a community-based intelligence-sharing mechanism ("Convey").
Co-Authors Ade Surya Bakti Pane Astri Mutia Rahma Erbin Sitorus Khairul Rian Farta Wijaya Sitorus, Zulham