cover
Contact Name
Andry Fajar Zulkarnain
Contact Email
andry.zulkarnain@ulm.ac.id
Phone
+6281223932020
Journal Mail Official
andry.zulkarnain@ulm.ac.id
Editorial Address
Jl. Brigjen H. Hasan Basry Komp. Kampus ULM Kayu Tangi Banjarmasin, Kalimantan Selatan Phone / Fax: 0511-3304405
Location
Kota banjarmasin,
Kalimantan selatan
INDONESIA
JTIULM (Jurnal Teknologi Informasi Universitas Lambung Mangkurat)
ISSN : 25275399     EISSN : 25282514     DOI : http://dx.doi.org/10.20527
Jurnal Teknologi Informasi Universitas Lambung Mangkurat (JTIULM) is intended as a media for scientific studies on the results of research, thinking and analytical-critical studies regarding research in Systems Engineering, Informatics / Information Technology, Information Management and Information Systems. As part of the spirit of disseminating knowledge from the results of research and thought for service to the wider community and as a reference source for academics in the field of Technology and Information.
Articles 162 Documents
Edge AI Using MobileNet Architecture for Driver Drowsiness Detection Rafie, Rafi e
Jurnal Teknologi Informasi Universitas Lambung Mangkurat (JTIULM) Vol. 11 No. 1 (2026)
Publisher : Fakultas Teknik Universitas Lambung Mangkurat

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.20527/jtiulm.v11i1.517

Abstract

Driving safety is a crucial issue significantly influenced by the driver's physical condition, where fatigue and drowsiness are major factors causing traffic accidents. This study aims to develop a real-time drowsiness detection system utilizing Edge AI technology based on the MobileNet architecture. This architecture was selected due to its efficiency in performing image classification on resource-constrained devices. The dataset used consists of 4,000 digital images balanced into open-eye and closed-eye classes. The model was trained using the TensorFlow framework and optimized through post-training quantization into the TensorFlow Lite format to reduce model size and inference latency. Performance evaluation was conducted by testing 372 new test images. The results indicate that the balanced model achieved an accuracy rate of 94%. Confusion matrix analysis showed a precision value of 1.000 for the closed-eye class and a recall of 1.000 for the open-eye class, indicating that the system is highly reliable in minimizing detection errors. With processing speeds reaching 10 to 22 Frames Per Second (FPS) on edge devices, this system is proven effective for implementation as a responsive driving safety assistant. Drowsiness detection duration indicator “Closed: 0.32s” represents part of the system logic used to trigger an alert. The system does not immediately activate an alarm during normal blinking, it measures the duration of eye closure. If the duration exceeds a predefined threshold (e.g., >0.30 seconds), an alert is triggered in the form of an audible alarm
Hybrid CNN Feature Extraction and Machine Learning Classification for UAV-Based Vegetation Density Land Cover Mapping in Peatlands Maulidiya, Erika; Sari, Yuslena; Gani, Irham Maulani Abdul; Islami, Achmad Mujaddid; Yunita, Helda
Jurnal Teknologi Informasi Universitas Lambung Mangkurat (JTIULM) Vol. 11 No. 1 (2026)
Publisher : Fakultas Teknik Universitas Lambung Mangkurat

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.20527/jtiulm.v11i1.518

Abstract

Land cover classification from UAV imagery has become increasingly important for environmental monitoring, especially in peatland ecosystems where vegetation density is closely related to land openness, degradation risk, and fire vulnerability. This study proposes a hybrid approach that combines CNN-based feature extraction with machine learning classification for vegetation density-based land cover classification. A total of 3,000 UAV images collected from Block 1 of the Liang Anggang Protected Forest, Banjarbaru, were used in this study and categorized into three classes: bare, moderate, and high vegetation density. The images were preprocessed through cropping, resizing, and labeling prior to feature extraction. ResNet-50 and DenseNet-121 were employed as feature extractors, while ten machine learning classifiers were evaluated, namely CalibratedClassifierCV, SVC, NuSVC, LogisticRegression, PassiveAggressiveClassifier, SGDClassifier, LinearSVC, XGBClassifier, Perceptron, and LGBMClassifier. The results show that ResNet-50 generally outperformed DenseNet-121 as a feature extractor. The best and most balanced performance was achieved by the ResNet-50 + SVC combination, which obtained 84% accuracy, 84% F1-score, 91% precision, 77% recall, and a computation time of 8.58 minutes. Although CalibratedClassifierCV achieved the same accuracy, it required substantially longer processing time. These findings indicate that classification performance is influenced not only by the classifier used, but also by the compatibility between feature representation and classification mechanism. Therefore, the combination of ResNet-50 and SVC is recommended for UAV-based vegetation density land cover classification in peatlands.
Web-Based Linear Regression Algorithm for Used Motorcycle Sales Prediction to Maintain Stability and Improve Sales at Alifa Jaya Motor Showroom Muhamad Asykal Rizki; Dola Irwanto
Jurnal Teknologi Informasi Universitas Lambung Mangkurat (JTIULM) Vol. 11 No. 2 (2026)
Publisher : Fakultas Teknik Universitas Lambung Mangkurat

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.20527/jtiulm.v11i2.524

Abstract

The growth of the automotive industry in Indonesia, particularly in the used motorcycle sales sector, demands more accurate and analytics-driven data management to support operational decision-making. Alifa Jaya Motor Showroom, located in Subang, still manages sales data manually, making it difficult to predict sales trends and optimally regulate stock availability. An imbalance between inventory supply and market demand risks creating unit surpluses or shortages that negatively impact the showroom's operational stability. This study aims to design and implement a web-based used motorcycle sales prediction system using a simple linear regression algorithm. System development employs the Prototype method, carried out iteratively through requirements gathering, design, prototype construction, evaluation, and refinement stages. The built system encompasses sales and motorcycle data management, linear regression-based prediction processing, sales trend graph visualization, and presentation of statistical parameters including the regression equation, coefficient of determination (R²), and Mean Absolute Percentage Error (MAPE), supported by PHP, MySQL, HTML, CSS, and Bootstrap technologies. Black Box testing results confirm that all features function according to specified requirements. Evaluation across 12 historical data periods yields an average MAPE of 23,13% (fair category) with an R² value of 0,0806, indicating that the model remains feasible as a decision-support basis, even though the time variable does not yet fully explain sales variation. Implementation of this system is expected to help Alifa Jaya Motor Showroom improve data management effectiveness, reduce stock imbalance risks, and maintain overall sales stability and volume.
Development of a Desktop-Based Digital Document Compression and Decompression Application Using Huffman and Lempel–Ziv–Welch (LZW) Algorithms with Compression Ratio Evaluation Eko Purwadi; Sri Rama Putri
Jurnal Teknologi Informasi Universitas Lambung Mangkurat (JTIULM) Vol. 11 No. 2 (2026)
Publisher : Fakultas Teknik Universitas Lambung Mangkurat

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.20527/jtiulm.v11i2.525

Abstract

The rapid growth of digital documents has increased the demand for storage capacity and efficient data distribution. Consequently, robust data compression techniques have become essential to mitigate storage constraints by reducing file sizes without compromising data integrity. This condition highlights the importance of data compression techniques capable of reducing file size without losing information. This study aims to develop a desktop-based application for digital document compression and decompression using the Huffman and Lempel–Ziv–Welch (LZW) algorithms while evaluating their performance based on compression ratio, space saving, and processing time. The application was developed using the Waterfall model with the Python programming language and the PyQt6 framework. System testing consisted of Black Box Testing to verify application functionality and White Box Testing using the Basis Path Testing technique to evaluate the program logic. Performance evaluation was conducted on five document formats, namely .txt, .pdf, .docx, .xlsx, and .pptx. The results indicate that the application successfully performs document compression and decompression across all tested document formats. Functional testing confirmed that all application features operated as expected, while the performance evaluation showed that the Huffman algorithm provided more stable compression performance across different document types. The performance evaluation demonstrated that the Huffman algorithm yielded a stable, albeit moderate, compression performance across different document types. In contrast, the Lempel–Ziv–Welch (LZW) algorithm achieved better compression results for text documents containing repetitive character patterns. These findings indicate that the effectiveness of a compression algorithm is strongly influenced by the characteristics of the processed data; therefore, selecting an appropriate algorithm should consider the document type.
Implementation of Autoencoder Model Based on Deep Learning for Keylogger Activity Detection on Android Through System Anomaly Analysis Naufal Fauzi Akbar; Saprudin
Jurnal Teknologi Informasi Universitas Lambung Mangkurat (JTIULM) Vol. 11 No. 2 (2026)
Publisher : Fakultas Teknik Universitas Lambung Mangkurat

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.20527/jtiulm.v11i2.527

Abstract

The increasingly widespread use of Android devices has heightened the risk to personal data security, particularly through keylogger activities capable of covertly intercepting user keystrokes without being detected by conventional signature-based security systems. This study aims to design and implement a keylogger activity detection system on Android devices using a deep learning-based Autoencoder model through a system anomaly analysis approach. A quantitative experimental method with an unsupervised anomaly detection approach was employed, in which the model was trained exclusively on normal device condition data. Data were collected through direct observation and internal logging mechanisms implemented in the developed KeyGuard application, yielding 904 data samples comprising CPU usage, RAM usage, network traffic, Accessibility Status, and Accessibility Events parameters, consisting of 764 normal-condition samples and 140 anomaly-condition samples obtained through testing with the Tispy keylogger application. The dataset was partitioned into training and testing sets, with evaluation conducted on 292 testing samples using a Confusion Matrix. The results indicate that the Autoencoder model, with a threshold value of 0.36, achieved an Accuracy of 96.57%, Precision of 95.77%, Recall of 97.14%, and F1-Score of 96.45%. The system successfully detected 136 of 140 tested keylogger activities, with a low false positive rate of six samples. This study concludes that the Autoencoder-based system anomaly analysis approach is effective for detecting keylogger activity on Android devices without relying on a signature database, thereby contributing to the development of more adaptive mobile security solutions capable of identifying previously unrecognized threats.
Evaluation and Gap Mapping of the SPBE Management Domain at BAPPERIDA Pekalongan City Using the PermenPAN-RB No. 59/2020 Framework Mukti Ahmad Sholehudin; Ahsanul Fikri; Aji Supriyanto
Jurnal Teknologi Informasi Universitas Lambung Mangkurat (JTIULM) Vol. 11 No. 2 (2026)
Publisher : Fakultas Teknik Universitas Lambung Mangkurat

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.20527/jtiulm.v11i2.528

Abstract

The Electronic-Based Government System (SPBE) in technical planning agencies requires systematic, evidence-based measurement. This study evaluates the SPBE Management Domain maturity at BAPPERIDA Pekalongan City using the PermenPAN-RB No. 59/2020 framework. Unlike perception-based studies, this research applies an evidence-based document review method confirmed through semi-structured interviews, covering six management aspects: risk management, information security, data management, ICT asset management, HR management, and change management, with 25 supporting documents examined. Results show that BAPPERIDA is at Level 2 (Developing) with an average index of 2.29 out of 5.00. The highest-scoring aspects are information security, ICT asset management, and HR management (2.50), while risk management and data management score lowest (2.00). The primary gap is the absence of key operational documents such as Risk Registers, data catalogs, and change management policies. This study produces a measurable gap map and a three-phase improvement roadmap toward Level 3, contributing an evidence-based OPD-level assessment model replicable across regional planning agencies.
Analysis and Redesign of the Social Media Content Production Business Process Using Value Stream Mapping (VSM) at SMK Muhammadiyah 5 Kepanjen Wahyu Candra Setyaji; Novia Ratnasari; Farid Wahyudi; Anisa Hudi Widaningrum; Muhammad Hanif Fahmi
Jurnal Teknologi Informasi Universitas Lambung Mangkurat (JTIULM) Vol. 11 No. 2 (2026)
Publisher : Fakultas Teknik Universitas Lambung Mangkurat

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.20527/jtiulm.v11i2.529

Abstract

Social media content production at SMK Muhammadiyah 5 Kepanjen faces operational inefficiency issues characterized by high waiting times and media team overburden due to a lengthy review bureaucracy. Although the Value Stream Mapping (VSM) method is commonly used in the manufacturing sector, its adaptation to map non-physical workflow waste (information waste) in digital content production within vocational education institutions remains scarce (research gap). Therefore, this study provides novelty and scientific contribution by adapting the Lean Service approach through VSM to redesign the creative industry business process in an educational environment. This research is qualitative, utilizing a case study approach. Data were collected through observation (time study), in-depth interviews, and documentation, validated using technical and source triangulation. The Current State mapping revealed a low Process Cycle Efficiency of 44.74% with a total Lead Time of 406.25 minutes. Fishbone Diagram analysis found that the main bottlenecks stem from the waste of waiting and waste of defect caused by the absence of written preliminary instructions and a centralized review authority. Through the Future State design and the proposed refinement of the Standard Operating Procedure (SOP), which include standardized written briefs and task delegation, the estimated total Lead Time was successfully reduced by 164.5 minutes, projecting a significant PCE increase to 75.18%. Practically, the results of this study can serve as a reference for other schools or institutions in improving the efficiency of their digital content production processes.
Implementation of the K-Nearest Neighbor (KNN) Algorithm for Website-Based Coffee Recommendations Based on Customer Preference Profiles (Case Study: 5CM Coffee and Eatery) Satrio Miftachul Choiri; Achmad Nur Sholeh
Jurnal Teknologi Informasi Universitas Lambung Mangkurat (JTIULM) Vol. 11 No. 2 (2026)
Publisher : Fakultas Teknik Universitas Lambung Mangkurat

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.20527/jtiulm.v11i2.530

Abstract

Menu selection at 5CM Coffee and Eatery is still carried out manually by the barista, so signature drink suggestions vary between staff and become inconsistent during busy hours, especially as the growing menu variety leads to customer choice overload. This study implements a website-based coffee recommendation system using the K-Nearest Neighbor (KNN) algorithm to address this problem. Nine beverage attributes—sweetness, bitterness, acidity, body, aroma, milk content, serving temperature, coffee type, and caffeine level—rated by the barista serve as training data, while customer preferences entered through the website serve as testing data; both are normalized with MinMaxScaler and compared using Euclidean Distance at k = 3. The system was evaluated through White Box Testing, Black Box Testing, and a User Acceptance Test. White Box Testing on three core modules produced Cyclomatic Complexity values of 3, 5, and 6 with all logical paths verified, Black Box Testing on 24 scenarios ran without error, and the UAT of 301 respondents recorded an overall acceptance score of 77.97%, categorized as Good. These results show that the implemented KNN algorithm produces relevant, personalized, and well-accepted signature drink recommendations, offering a more consistent alternative to manual recommendations at 5CM Coffee and Eatery.
Application of Rule-Based and Threshold-Based Methods in Determining Administrative Archive Status Jidan Juliana; Marsani Asfi; Mesi Febima
Jurnal Teknologi Informasi Universitas Lambung Mangkurat (JTIULM) Vol. 11 No. 2 (2026)
Publisher : Fakultas Teknik Universitas Lambung Mangkurat

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.20527/jtiulm.v11i2.536

Abstract

Balai Besar Wilayah Sungai (BBWS) Cimanuk Cisanggarung in Cirebon City faced challenges in determining the status of administrative archives because the classification of archive type, category, and status was still performed manually, making the process prone to inconsistency and time-consuming. This study aimed to design and build an information system applying the Rule-Based System and Threshold-Based methods to support automatic determination of archive status. Research data consisted of administrative archive metadata of BBWS Cimanuk Cisanggarung for the 2022-2025 period, obtained through observation, interviews, and literature study. The Rule-Based System applied IF-THEN rules to determine archive type and category based on the Records Retention Schedule (Jadwal Retensi Arsip/JRA) and determined the initial status based on archive age, producing 349 Active archives and 151 Inactive archives. The Inactive archives were then evaluated using the Threshold-Based System, which calculated the total score from the sum of the archive type score (Sj) and the archive age score (Su), resulting in 5 Permanent archives, 6 Evaluation archives, and 140 Destroyed archives. The system was implemented using the PHP programming language with the Laravel framework and a MySQL database, and was tested using the Black Box Testing method. The test results showed that all system features functioned according to requirements, the Rule-Based System classification results were consistent with the JRA, and the Threshold-Based System results were consistent with the archivist's manual decisions. Thus, the developed system helped BBWS Cimanuk Cisanggarung determine the status of administrative archives more quickly, consistently, and with proper documentation compared to the previous manual process.
Developing A Student Dropout Detection System By Using The KNN Model Muhammad Hafriyandi Saputra; Pitrasacha Adytia; Wahyuni Wahyuni
Jurnal Teknologi Informasi Universitas Lambung Mangkurat (JTIULM) Vol. 11 No. 2 (2026)
Publisher : Fakultas Teknik Universitas Lambung Mangkurat

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.20527/jtiulm.v11i2.537

Abstract

The high number of students experiencing dropout is a serious challenge that must be addressed by higher education institutions, including STMIK Widya Cipta Dharma. This study aims to propose an early detection system that demonstrates the ability to identify students who are at risk of dropping out based on their academic data. The system is developed using the K-Nearest Neighbor (KNN) algorithm as a classification method to predict the risk of student dropout. The data used includes Grade Point Average (GPA), total credits earned, and the duration of study. The test results show that the KNN algorithm can provide a high level of accuracy in categorizing students into dropout risk levels. This system can serve as a support tool for the institution in determining preventive actions to reduce the number of students at risk of failing to complete their studies and to improve the quality of academic guidance.