cover
Contact Name
Muhammad Wali
Contact Email
muhammadwali487@gmail.com
Phone
+6285277777449
Journal Mail Official
cj@ypmma.org
Editorial Address
Jl. B. Aceh – Medan GP. Pasir Putih Kec. Peureulak Aceh Timur.
Location
Kab. aceh timur,
Aceh
INDONESIA
Computer Journal
Published by Yayasan YPMMA
ISSN : 29646200     EISSN : 29646219     DOI : https://doi.org/10.58477/cj
Computer Journal, e-ISSN: 2964-6219 and p-ISSN: 2964-6200 is a free and open-access journal published by the Research Division, YPMMA Institute, Indonesia. Computer Journal is an international, scientific, peer-reviewed, open-access computer science journal, including computer and network architecture and human-computer interaction as its main focus, published six months online. Computers Journal is an international, open access journal which provides an advanced forum for computer sciences. It publishes reviews, regular research papers and short communications. Our aim is to encourage scientists to publish their experimental and theoretical results in as much detail as possible. There is no restriction on the length of the papers. The full experimental details must be provided so that the results can be reproduced.
Articles 98 Documents
Comparison of Support Vector Machine and K-Nearest Neighbor Methods for Predicting Student Learning Outcomes Based on Student Performance Data Chynthia Drinita; Felicia Felicia; Palma Juanta
Computer Journal Vol. 4 No. 2 (2026): August
Publisher : Yayasan Pendidikan Mitra Mandiri Aceh (YPMMA)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.58477/cj.v4i2.512

Abstract

Information systems and machine learning have encouraged the use of academic data to predict student learning outcomes in the education sector. This study evaluated the performance of Support Vector Machine (SVM) and K-Nearest Neighbor (KNN) algorithms in predicting student learning outcomes using the Student Performance in Exams Dataset from Kaggle. A quantitative computational experiment was conducted through several preprocessing stages, including target variable creation, categorical data conversion, feature standardization using StandardScaler, and data splitting into 80% training data and 20% testing data. Model performance was evaluated using accuracy, precision, recall, F1-score, Receiver Operating Characteristic (ROC) curve, and Area Under the Curve (AUC). The results show that SVM achieved an accuracy of 97%, with average precision, recall, and F1-score values of 0.97, while KNN achieved an accuracy of 90% and an F1-score of 0.88. The AUC values of 1.00 for SVM and 0.97 for KNN indicate that both models performed well, although SVM provided better class separation. Therefore, SVM was more effective than KNN in predicting student learning outcomes on the dataset used and may support the development of machine learning-based academic prediction systems.
Evaluasi Usability Mentari UNPAM Menggunakan Model SUS (System Usability Scale) Firman Pratama; Devi Damayanti; Haekal Mimtazulfaqhi Zaydan; Gilang Ramadhan; M. Hafizh Rafid Hidayat
Computer Journal Vol. 4 No. 2 (2026): August
Publisher : Yayasan Pendidikan Mitra Mandiri Aceh (YPMMA)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.58477/cj.v4i2.513

Abstract

Current higher education infrastructure heavily relies on Learning Management Systems (LMS) as digital hubs for academic interaction. Universitas Pamulang implemented Mentari UNPAM to support large-scale online learning. However, the platform’s usability and User Experience (UX) have not been empirically assessed using standardized metrics. This lack of evidence contributes to increased cognitive load for users and limits the implementation of User-Centered Design (UCD). This study aims to evaluate the usability level of Mentari UNPAM and propose interface redesign recommendations. A descriptive quantitative approach was conducted by administering the System Usability Scale (SUS) to 150 active Informatics Engineering students. The results show an average SUS score of 68.27. Based on the Sauro-Lewis scale, this corresponds to a Fair-to-Good range with an Acceptable rating. Data distribution analysis further indicates satisfaction polarization associated with interaction friction during the First-Time User Experience (FTUE) phase, navigation disorientation, and server latency during peak traffic. As mitigation, the study recommends restructuring the layout hierarchy of critical features, integrating breadcrumb navigation, providing digital onboarding, and optimizing server architecture.
Implementasi Machine Learning untuk Deteksi Kejujuran Wakif pada Sistem Wakaf Digital Berbasis Blockchain Rahma Nadiya
Computer Journal Vol. 5 No. 1 (2027): February
Publisher : Yayasan Pendidikan Mitra Mandiri Aceh (YPMMA)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.58477/cj.v5i1.507

Abstract

Digital waqf requires a validation mechanism that not only protects data integrity after storage but also evaluates transaction credibility before the data becomes immutable on a blockchain. This study develops an intelligent pre-validation mechanism through a hybrid architecture integrating a web application, machine learning, a backend API, smart contracts, and blockchain. The dataset combined the Ethereum Fraud Detection Dataset (9,841 transactions and 51 attributes) with transaction and behavioral data from the digital waqf system. Data processing included cleaning, missing-value handling, categorical encoding, StandardScaler normalization, feature engineering, and Isolation Forest anomaly detection. The resulting anomaly score was normalized into a 0–100 trust score and passed to the smart contract to process high-trust transactions, route medium-risk transactions to nazhir review, and block high-risk transactions before on-chain recording. Built using Python, Scikit-learn, PHP, MySQL, Solidity, Ganache, and MetaMask, the prototype demonstrated an end-to-end pre-validation flow from waqif input to on-chain decision-making. Representative tests yielded trust scores of 92 (normal), 68 (review), and 27 (fraud), showing that separating off-chain inference from on-chain validation provides an effective control layer against garbage in, garbage out without imposing machine learning computation on the smart contract.
Model Explainable Machine Learning untuk Deteksi Anomali Transaksi Blockchain Bitcoin Menggunakan Analisis SHAP Muhammad Ibnu Fauzan
Computer Journal Vol. 5 No. 1 (2027): February
Publisher : Yayasan Pendidikan Mitra Mandiri Aceh (YPMMA)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.58477/cj.v5i1.514

Abstract

The growing volume of Bitcoin transactions requires anomaly detection mechanisms that are both accurate and explainable, as high-performing classifiers often lack decision traceability for auditors and security analysts. This study develops an explainable machine learning approach for Bitcoin transaction anomaly detection by comparing Logistic Regression, Random Forest, and Extreme Gradient Boosting (XGBoost), followed by SHapley Additive exPlanations (SHAP) interpretation on the best model. Using the Elliptic Bitcoin Dataset, preprocessing included feature-label merging, removing unknown labels, binary target conversion, selecting 165 numerical features, MinMaxScaler normalization, and an 80:20 stratified train-test split, retaining 14,253 labeled transactions: 13,380 licit (93.88%) and 873 illicit (6.12%). Evaluation based on accuracy, precision, recall, F1-score, ROC-AUC, confusion matrix, and training time showed that XGBoost achieved the best performance with 0.9961 accuracy, 1.0000 precision, 0.9371 recall, 0.9676 F1-score, and 0.9960 ROC-AUC, yielding 2,676 true negatives, zero false positives, 164 true positives, and 11 false negatives. Transaction-level SHAP analysis revealed how individual feature contributions shifted predictions toward the licit or illicit class, demonstrating that integrating XGBoost and SHAP delivers accurate anomaly detection with transparent decision support for blockchain investigations.
Pemetaan Wilayah di Kota Palembang untuk Memprioritaskan Penyaluran Bantuan Sosial Menggunakan Algoritma K-Means Clustering M. Syahril M. Syahril; Devi Udariansyah; Alex Wijaya
Computer Journal Vol. 5 No. 1 (2027): February
Publisher : Yayasan Pendidikan Mitra Mandiri Aceh (YPMMA)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.58477/cj.v5i1.520

Abstract

Poverty remains a multidimensional social issue in Palembang City, where the poor population reached 10.22% in 2023. The distribution of social assistance often faces challenges such as data inaccuracies, uneven distribution, and subjective decision-making processes. This study aims to apply the K-Means Clustering algorithm to map 18 sub-districts in Palembang City into priority zones for social assistance distribution based on seven socio-economic indicators. The dataset includes poor population, unemployment rate, average income, population density, public facility access score, number of social assistance recipient households, and the Human Development Index (HDI). Data preprocessing was performed using Min-Max Normalization, followed by the calculation of the Composite Vulnerability Index (CVI). Model evaluation using the Elbow Method, Silhouette Score (0.45–0.55), and Davies–Bouldin Index confirmed K=3 as the optimal number of clusters. The resulting clusters categorize the sub-districts into high priority (emergency, e.g., Kertapati, Gandus, and Seberang Ulu I), medium priority (alert, e.g., Plaju, Sukarami, and Sako), and low priority (independent, e.g., Ilir Timur I, Bukit Kecil, and Alang-Alang Lebar). In addition, a Decision Support System (DSS) simulator was implemented based on the clustering and CVI results to provide data-driven budget and quota allocation recommendations.
Perceived Accuracy and Relevance of AI-Based Product Recommendations and Generation X Purchase Decisions in Jakarta Dita Madonna Simanjuntak; Amri Sinuraya; Selvia De Wana
Computer Journal Vol. 5 No. 1 (2027): February
Publisher : Yayasan Pendidikan Mitra Mandiri Aceh (YPMMA)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.58477/cj.v5i1.528

Abstract

This study examines how perceived accuracy and contextual relevance of artificial-intelligence-based product recommendations influence purchase decisions among Generation X e-commerce users in Jakarta. A quantitative explanatory design was modeled with 300 respondent profiles and 18 reflective indicators. Partial least squares structural equation modeling, 10,000-resample bootstrapping, coefficient-difference testing, and tenfold cross-validation were applied. Perceived accuracy had a stronger positive effect on purchase decisions (β=.604, p<.001) than relevance (β=.217, p<.001), and the difference was significant (Δβ=.387, p<.001). The model explained 54.2% of purchase-decision variance and showed positive Qpredict2​ values. The findings indicate an accuracy–relevance asymmetry: relevance attracts attention, whereas perceived accuracy more strongly reinforces purchase decisions.
Artificial Intelligence-Based Spatiotemporal Clustering for Identifying Homogeneous Climatic Zones in West Java Oktaviana Indriani; Eka Ardhianto
Computer Journal Vol. 5 No. 1 (2027): February
Publisher : Yayasan Pendidikan Mitra Mandiri Aceh (YPMMA)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.58477/cj.v5i1.542

Abstract

Homogeneous Climatic Zones (ZOM) support operational seasonal services in Indonesia, particularly across regions where rainfall changes sharply over short distances. This study evaluates a climatic regionalization framework that combines Uniform Manifold Approximation and Projection (UMAP) and Hierarchical Density-Based Spatial Clustering of Applications with Noise (HDBSCAN) using BMKG gridded rainfall data across West Java at approximately 5 km resolution for the 1991–2020 period. Each of the 1,214 grid cells was represented by a 36-dasarian climatological rainfall profile. A search across 360 hyperparameter combinations yielded 324 executable configurations, followed by spatial absorption of transition grids and layered evaluation using the Silhouette Score, Heterogeneity Index (HI), and operational tractability. Scenario 60 (n_neighbors = 5, min_dist = 0.01, min_cluster_size = 10, min_samples = 5) produced 54 initial zones with a Silhouette Score of 0.5633 and an HI of 7.239, compared with an HI of 8.464 for the 41-zone operational baseline. Subsequent spatial refinement produced a corrected final map containing 50 sequential ZOM. Distinguishing the evaluation metrics of the 54-zone candidate stage from the 50-zone cartographic structure of the spatial refinement stage clarifies the analytical sequence and supports a more precise interpretation of cluster quality, climatic homogeneity, and final zone boundaries.
Model Sistem Pendukung Keputusan untuk Penilaian Risiko Stunting pada Balita Menggunakan Certainty Factor Berbasis Bukti: Studi Pembuktian Konsep Muhammad Nurwegiono; Paulus Lucky Tirma Irawan; Ummu Ditya Erliana; Mahendra Data
Computer Journal Vol. 5 No. 1 (2027): February
Publisher : Yayasan Pendidikan Mitra Mandiri Aceh (YPMMA)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.58477/cj.v5i1.567

Abstract

Childhood stunting is multifactorial, and its risk assessment often involves interdependent evidence. This study developed a proof-of-concept decision support model based on the Certainty Factor (CF) method using literature-derived evidence weights. Fifteen risk factors were curated from 22 District Health Office indicators. Odds ratios were transformed into attributable fractions and mapped to provisional evidence weights using explicit rounding, capping, and attenuation rules. To reduce double counting, the hierarchical CF model groups the factors into seven clusters and applies a within-cluster correlation factor (β) alongside a mediation factor (α) for selected upstream clusters. The web prototype was computationally trace-tested using five stunting cases, which were not used for clinical or diagnostic validation. Under the baseline parameter setting, hierarchical CF values were 0.027–0.113 lower than classical CF values, and sensitivity analysis across scenarios ranging from strong to no correction showed no category changes across the five traces. The prototype accurately reproduced manual calculations and preserved factor-level traceability. Expert elicitation, threshold calibration, and diagnostic evaluation using representative case-control data remain necessary before field implementation.

Page 10 of 10 | Total Record : 98