Article Per Year (5 Year)
p-Index From 2021 - 2026
0.444
P-Index
This Author published in this journals
All Journal Jurnal Nasional Komputasi dan Teknologi Informasi
Fadilla Amanah
Universitas Negeri Medan
Aerospace Engineering Automotive Engineering Computer Science & IT Control & Systems Engineering Decision Sciences, Operations Research & Management Electrical & Electronics Engineering Engineering Neuroscience Transportation

Published : 2 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 2 Documents
Search

 1 
Pengembangan Sistem Deteksi Malware Berbasis Web Menggunakan Algoritma Random Forest dengan Notifikasi Realtime Muhammad Naufal Musyaafa; Dedy Kiswanto; Ananda Irya Shakila Syukron; Fadilla Amanah

Publisher :

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.32672/jnkti.v8i6.9896

Abstract

Abstrak - Perkembangan teknologi digital yang pesat telah meningkatkan risiko serangan malware yang semakin kompleks dan sulit dideteksi dengan metode konvensional berbasis tanda tangan (signature-based). Penelitian ini bertujuan untuk mengembangkan sistem deteksi malware berbasis web menggunakan algoritma Random Forest dengan fitur notifikasi realtime untuk meningkatkan efektivitas dan kecepatan respon terhadap ancaman keamanan. Dataset yang digunakan berasal dari platform Kaggle, terdiri atas sekitar 34.000 file Portable Executable (PE) Windows dengan 55 fitur yang diekstraksi secara statis dari PE Header dan daftar impor DLL. Model machine learning dibangun menggunakan Pipeline dari Scikit-learn yang terdiri dari StandardScaler, CountVectorizer, dan RandomForestClassifier. Hasil evaluasi menunjukkan bahwa model mencapai tingkat akurasi sebesar 99,65% dengan nilai precision, recall, dan F1-score sempurna (1.00) pada kedua kelas, menunjukkan kemampuan model yang sangat andal dan seimbang dalam mendeteksi file berbahaya maupun aman. Model yang telah divalidasi kemudian diimplementasikan ke dalam aplikasi web berbasis Django dengan antarmuka dinamis, yang memberikan peringatan instan kepada pengguna secara asinkronus menggunakan teknologi JavaScript (Fetch API). Penelitian ini berhasil menghasilkan sistem deteksi malware yang praktis, adaptif, dan responsif, yang dapat meningkatkan kesadaran serta keamanan digital pengguna terhadap ancaman siber yang terus berkembang.Kata kunci : Deteksi Malware; Machine Learning; Random Forest; Django; Notifikasi Realtime; Abstract - The rapid development of digital technology has increased the risk of malware attacks that are increasingly complex and difficult to detect using conventional signature-based methods. This study aims to develop a web-based malware detection system using the Random Forest algorithm with real-time notification features to improve the effectiveness and speed of response to security threats. The dataset used was sourced from the Kaggle platform, consisting of approximately 34,000 Windows Portable Executable (PE) files with 55 features extracted statically from the PE Header and DLL import list. The machine learning model was built using a Pipeline from Scikit-learn consisting of StandardScaler, CountVectorizer, and RandomForestClassifier. The evaluation results show that the model achieved an accuracy rate of 99.65% with perfect precision, recall, and F1-score (1.00) in both classes, demonstrating the model's highly reliable and balanced ability to detect both malicious and safe files. The validated model was then implemented into a Django-based web application with a dynamic, which provides instant alerts to users asynchronously using JavaScript technology (Fetch API).This research has successfully produced a practical, adaptive, and responsive malware detection system that can increase user awareness and digital security against ever-evolving cyber threats. Keywords: Malware Detection; Machine Learning; Random Forest; Django; Real-Time Notifications;
Sistem Keamanan Pintu Berbasis Computer Vision dengan Biometric Face Recognition dan Physical Tampering Detection Felix John Pardamean Hutabarat; Dedy Kiswanto; Paskah Abadi Simanullang; Fadilla Amanah

Publisher :

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.32672/jnkti.v9i1.10270

Abstract

Abstrak - Keamanan akses pintu pada lingkungan hunian dan kos membutuhkan sistem yang tidak hanya mampu memverifikasi identitas pengguna, tetapi juga responsif terhadap ancaman fisik terhadap perangkat. Penelitian ini mengusulkan dan merealisasikan S.I.G.H.T. (Secure Intelligent Gate Hardware Tamper-detection), sebuah sistem keamanan pintu berbasis computer vision yang menggabungkan biometric face recognition menggunakan algoritma Local Binary Patterns Histograms (LBPH) dengan physical tampering detection berbasis sensor getaran. Arsitektur sistem terdiri atas backend FastAPI, dashboard web berbasis React sebagai pusat pemantauan dan kontrol, agen kamera untuk pemrosesan citra pada perangkat edge, serta modul IoT gerbang yang mengendalikan kunci dan alarm secara real-time melalui antrian perintah terpusat. Metode pengembangan yang digunakan adalah pendekatan Research and Development (RD) dengan model Waterfall yang mencakup analisis kebutuhan, perancangan, implementasi, dan pengujian terstruktur. Validasi fungsional dilakukan menggunakan black box testing pada skenario utama, seperti otentikasi wajah, pengelolaan data penghuni, respons sensor getaran, dan kontrol aktuator pintu serta alarm. Hasil pengujian menunjukkan seluruh skenario berjalan sesuai harapan dengan status “Lulus”, sehingga S.I.G.H.T. dinilai layak sebagai prototipe solusi keamanan pintu berlapis yang adaptif dan berpotensi dikembangkan lebih lanjut pada skala implementasi yang lebih luas.Kata kunci : Computer vision; Deteksi tampering fisik; Internet of Things; LBPH; Sistem keamanan pintu; Abstract - Door access security in residential and boarding environments requires a system that not only verifies user identity, but also responds to physical threats directed at the device. This study proposes and implements S.I.G.H.T. (Secure Intelligent Gate Hardware Tamper-detection), a computer-vision-based door security system that combines biometric face recognition using the Local Binary Patterns Histograms (LBPH) algorithm with physical tampering detection using a vibration sensor. The system architecture consists of a FastAPI backend, a React-based web dashboard as the central monitoring and control interface, a camera agent for image processing on edge devices, and an IoT gate module that controls the lock and alarm in real time through a centralized command queue. The development process follows a Research and Development (RD) approach with the Waterfall model, covering requirements analysis, system design, implementation, and structured testing stages. Functional validation is carried out using black box testing on key scenarios such as face authentication, resident data management, vibration sensor response, and actuator control for the door and alarm. The results show that all scenarios meet the expected outcomes with a “Pass” status, indicating that S.I.G.H.T. is feasible as a layered and adaptive door security prototype that can be further extended to broader deployment contexts.Keywords: Computer vision; Door security system; Internet of Things; LBPH; Physical tampering detection;
Co-Authors Ananda Irya Shakila Syukron Dedy Kiswanto Felix John Pardamean Hutabarat Muhammad Naufal Musyaafa Paskah Abadi Simanullang