cover
Contact Name
Ramdan Satra
Contact Email
Ramdan Satra
Phone
-
Journal Mail Official
ramdan@umi.ac.id
Editorial Address
-
Location
Kota makassar,
Sulawesi selatan
INDONESIA
ILKOM Jurnal Ilmiah
ISSN : 20871716     EISSN : 25487779     DOI : -
Core Subject : Science,
ILKOM Jurnal Ilmiah is an Indonesian scientific journal published by the Department of Information Technology, Faculty of Computer Science, Universitas Muslim Indonesia. ILKOM Jurnal Ilmiah covers all aspects of the latest outstanding research and developments in the field of Computer science, including Artificial intelligence, Computer architecture and engineering, Computer performance analysis, Computer graphics and visualization, Computer security and cryptography, Computational science, Computer networks, Concurrent, parallel and distributed systems, Databases, Human-computer interaction, Embedded system, and Software engineering.
Arjuna Subject : -
Articles 617 Documents
Comparative Study of Random Forest and Ordinal Regression in Concept Map Quality Assessment: The Role of TF-IDF, BERT, and SMOTE-based Balancing Rismayanti, Nurul; Prasetya, Didik Dwi; Widiyaningtyas, Triyanna; Hirashima, Tsukasa
ILKOM Jurnal Ilmiah Vol 17, No 3 (2025)
Publisher : Prodi Teknik Informatika FIK Universitas Muslim Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33096/ilkom.v17i3.2906.336-345

Abstract

Automatic assessment of concept map quality is an important challenge in the field of education, particularly in evaluating students' conceptual understanding objectively and efficiently. This study aims to compare the performance of two machine learning algorithms, namely Random Forest and Ordinal Regression, in classifying the quality of concept maps. The evaluation was conducted on three approaches to text feature representation: Term Frequency-Inverse Document Frequency (TF-IDF), Bidirectional Encoder Representations from Transformers (BERT), and a combination of both (TF-IDF + BERT). Additionally, this study compares the performance of the models under two dataset conditions: original data and data balanced using the Synthetic Minority Over-sampling Technique (SMOTE), to address the class imbalance that often occurs in educational data. The data used consists of a collection of propositions from students' concept maps that have been labeled with ordinal scores based on quality. Text representation is extracted using the TF-IDF and BERT approaches, and then used as input to build the classification model. Performance evaluation was conducted using the metrics of Accuracy, Precision, Recall, F1-score, Cohen’s Kappa, and MAE. The results show that the Ordinal Regression model with TF-IDF representation combined with SMOTE achieved the best performance, with an accuracy of 0.8777, an F1-score of 0.8773, and a Cohen’s Kappa of 0.7701. These results indicate that classical feature representations like TF-IDF remain effective in limited data scenarios, and that the SMOTE technique successfully improved the model's performance by reducing bias towards the majority class. This research contributes to the development of an automatic concept map assessment system and suggests optimal classification strategies for educational datasets with ordinal and imbalanced characteristics
LSTM-based Multivariate Time-Series Analysis: A Case of Journal Visitors Forecasting Saputra, Anggie Wahyu; Wibawa, Aji Prasetya; Pujianto, Utomo; Putra Utama, Agung Bella; Nafalski, Andrew
ILKOM Jurnal Ilmiah Vol 14, No 1 (2022)
Publisher : Prodi Teknik Informatika FIK Universitas Muslim Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33096/ilkom.v14i1.1106.57-62

Abstract

Forecasting is the process of predicting something in the future based on previous patterns. Forecasting will never be 100% accurate because the future has a problem of uncertainty. However, using the right method can make forecasting have a low error rate value to provide a good forecast for the future. This study aims to determine the effect of increasing the number of hidden layers and neurons on the performance of the long short-term memory (LSTM) forecasting method. LSTM performance measurement is done by root mean square error (RMSE) in various architectural scenarios. The LSTM algorithm is considered capable of handling long-term dependencies on its input and can predict data for a relatively long time. Based on research conducted from all models, the best results were obtained with an RMSE value of 0.699 obtained in model 1 with the number of hidden layers 2 and 64 neurons. Adding the number of hidden layers can significantly affect the RMSE results using neurons 16 and 32 in Model 1.
Enhancing Kubernetes-Based Microservices Deployment Efficiency Through DevOps and GitOps Maulana, Irvan; Umar, Rusydi; Yudhana, Anton
ILKOM Jurnal Ilmiah Vol 17, No 2 (2025)
Publisher : Prodi Teknik Informatika FIK Universitas Muslim Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33096/ilkom.v17i2.2562.107-119

Abstract

An effective and resilient means to deploy microservices to Kubernetes is an ongoing challenge. This challenge becomes more difficult with ever increasingly complex application architectures. This research explored a DevOps model based on GitOps that integrates ArgoCD and GitLab CI/CD, as a means to create a more effective, resilient, and scalable deployment. Twelve microservices that were deployed in a controlled experimentation format were used in a comparative approach to previous deployment practices that only considered manual deployments. The results show an overall deployment time improvement of 40%. For the deployments that were executed incorrectly, ArgoCD ensures service availability leveraging its self-healing capabilities. During the computation of each run we also experienced system performance in a sustained high-load environment. Upon high demand, we experienced the desired autoscaling behavior requested, which resulted in higher service responsiveness. In comparison to previous studies, this research considered statistical analysis, while also looking at an aspect of real-world orchestration and networking efficiency while adopting Kubernetes. Altogether, this research gives organizations practical advice on how they may optimize their deployment pipelines for efficient, scalable and resilient microservices.
Automated Hyperparameter Optimization of Lightweight YOLO11s for Efficient Road Crack Detection Angreni, Ida Ayu Ari; Diyanti, Diyanti; Valentine, Vega
ILKOM Jurnal Ilmiah Vol 18, No 1 (2026)
Publisher : Prodi Teknik Informatika FIK Universitas Muslim Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33096/ilkom.v18i1.2894.138-150

Abstract

Automatic road crack detection plays an essential role in infrastructure maintenance, where rapid and accurate visual inspection is required under real-world conditions. Although deep learning–based detection models have demonstrated promising performance, many existing approaches rely on computationally intensive architectures or require manual hyperparameter tuning, which limits their efficiency and real-time applicability. Moreover, the integration of lightweight detection models with automated hyperparameter optimization remains relatively underexplored.This study proposes an efficient road crack detection framework based on a lightweight YOLO11s architecture enhanced through automated hyperparameter optimization using Optuna on the DeepCrack dataset. The proposed methodology includes image preprocessing through data augmentation, normalization, and resizing to improve model robustness. Subsequently, key hyperparameters including learning rate, weight decay, dropout rate, and optimizer selection are automatically optimized to obtain the best model configuration. Experimental results indicate that the optimized YOLO11s model achieves a precision of 90.4%, recall of 86.8%, mAP@0.5 of 89.8%, and mAP@0.5:0.95 of 63.6% after 25 optimization trials. These results demonstrate that automated hyperparameter optimization can significantly improve detection performance while maintaining computational efficiency. The main contribution of this study lies in the systematic integration of automated hyperparameter tuning within a lightweight YOLO-based framework, providing a resource efficient and accurate solution suitable for real-time and large-scale road damage monitoring
SMOTE-Based Comparative Analysis of Machine Learning Models for Stroke Risk Prediction Using Imbalanced Healthcare Data Siregar, Ratu Mutiara; Satria, Budy; Fadilah, Sandi; Mayola, Liga; Safira, Silky
ILKOM Jurnal Ilmiah Vol 18, No 1 (2026)
Publisher : Prodi Teknik Informatika FIK Universitas Muslim Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33096/ilkom.v18i1.3161.180-194

Abstract

Stroke remains one of the leading causes of mortality and long-term disability worldwide, with a significant burden in Indonesia. Early detection is crucial, as up to 90% of stroke cases are potentially preventable through timely intervention. However, predictive modeling for stroke risk is often challenged by imbalanced datasets, where non-stroke cases significantly outnumber stroke cases, potentially biasing classification models. This study aims to perform a systematic comparative evaluation of six machine learning algorithms Logistic Regression, Decision Tree, Random Forest, Naïve Bayes, Support Vector Machine (SVM), and Extreme Gradient Boosting (XGBoost) for stroke risk prediction under imbalanced data conditions. The dataset consists of 5,110 patient records with 11 health-related features obtained from a publicly available healthcare dataset. Data preprocessing included anomaly removal, categorical encoding, feature scaling, and class balancing using the Synthetic Minority Oversampling Technique (SMOTE). Model evaluation was conducted using 5-fold cross-validation and assessed through accuracy, precision, recall, and F1-score metrics. The experimental results demonstrate that ensemble-based models outperform single classifiers. Random Forest achieved the highest mean accuracy of 97.12% (±0.42) with an F1-score of 0.96, followed closely by XGBoost with 96.85% (±0.51). Both models also exhibited superior recall performance, indicating improved minority class detection. The novelty of this study lies in the systematic evaluation of multiple machine learning models using SMOTE-based balancing and cross-validation on publicly available healthcare data, providing robust comparative insights for imbalanced medical classification problems.
Enhancing Eye Disease Classification Accuracy Using Convolutional Neural Networks with Transfer Learning Addyna, Nazlina Izmi; Sembiring, Rahmat Widia; Windarto, Agus Perdana
ILKOM Jurnal Ilmiah Vol 18, No 1 (2026)
Publisher : Prodi Teknik Informatika FIK Universitas Muslim Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33096/ilkom.v18i1.2886.195-206

Abstract

Eye diseases serve as a primary contributor to global blindness, making early detection a critical determinant in effective treatment outcomes. While retinal fundus image analysis is the diagnostic standard, conventional manual methods are often hindered by observer subjectivity and time inefficiencies. This study aims to optimize eye disease classification using a Convolutional Neural Network (CNN) approach empowered by transfer learning techniques. Utilizing a dataset of 1,200 retinal fundus images sourced from Kaggle, this research classifies four categories: normal, glaucoma, cataract, and diabetic retinopathy. To mitigate the challenge of limited labeled medical datasets, specific data augmentation strategies—including random flip, zoom, and contrast adjustments—were applied. The study conducts a comparative evaluation of three architectures: standard VGG16, baseline MobileNet, and a proposed optimized MobileNet. The proposed method utilizes Random Search to systematically optimize hyperparameters such as learning rates, dense layer units, and dropout rates. Experimental results demonstrate that the optimized MobileNet achieved superior performance with 89.17% accuracy, significantly outperforming the VGG16 baseline 82,00% and baseline MobileNet 85,00%. Notably, the model achieved perfect recall for diabetic retinopathy, although glaucoma remained the most challenging class due to subtle morphological similarities with normal eyes. These findings confirm that integrating lightweight CNNs with appropriate transfer learning yields a diagnostic system that is not only accurate but also efficient for deployment in resource-constrained environments
Smart Journal Finder: A Web-Based Scientific Article Categorization Using Jaccard Similarity Rodiah, Rodiah
ILKOM Jurnal Ilmiah Vol 18, No 1 (2026)
Publisher : Prodi Teknik Informatika FIK Universitas Muslim Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33096/ilkom.v18i1.2814.17-29

Abstract

The rapid growth of scientific publications presents challenges for researchers in identifying appropriate journals for manuscript submission. With an overwhelming number of journals across diverse disciplines, manually matching a manuscript to a suitable journal becomes inefficient and prone to misclassification. This study proposes the Smart Journal Finder, a web-based system designed to recommend relevant scientific journals by analyzing textual similarities between user-submitted manuscripts and indexed journal articles. The system processes input data including the title, abstract, keywords, and field of study through several stages: preprocessing, stop word removal, stemming using the Nazief-Adriani algorithm, and duplicate term elimination. Similarity scoring is performed using the Jaccard Similarity algorithm, followed by ranking the results and displaying journal metadata such as subject, publisher, and citation metrics. Results show that the system accurately transforms and filters input text, effectively calculates similarity scores, and successfully matches manuscripts to appropriate journals. By automating this process, the Smart Journal Finder enhances the efficiency of journal selection, improves the relevance of publication targets, and supports researchers in increasing the visibility and impact of their work. However, the current implementation is limited to Indonesian-language journals and does not yet incorporate semantic similarity or multilingual processing. Future work will focus on expanding coverage across disciplines and integrating more advanced similarity models.