Journal of Computer Networks, Architecture and High Performance Computing
Not a member yet
473 research outputs found
Sort by
Systematic Literature Review: A Comparison of Clustering Methods in Data Mining
No full textClustering is one of the fundamental techniques in data mining used to group data instances based on inherent similarities without relying on predefined labels. This technique plays a crucial role in numerous domains, including customer behavior analysis, pattern recognition, anomaly detection, bioinformatics, and many other applications that require a deeper understanding of hidden structures within data. Over the past decades, a wide range of clustering methods has been developed such as K-Means, DBSCAN, Hierarchical Clustering, density-based approaches, model-based clustering, and more recent algorithms that incorporate machine learning and deep learning paradigms. Each method offers distinct advantages and limitations and is suited for different data characteristics and analytical objectives. The SLR process includes identifying relevant articles, screening for quality and eligibility, extracting essential data, and synthesizing findings according to predefined systematic criteria. The primary aim of this review is to identify emerging research trends, understand methodological advancements, assess the performance of different clustering methods across diverse data contexts such as varying dataset sizes, noise levels, dimensionality, and cluster distributions and provide insights into the key factors that influence the selection of appropriate clustering techniques. The findings of this review indicate that no single clustering method consistently outperforms others in all scenarios. Certain algorithms may produce optimal results for low-dimensional datasets yet perform inadequately when applied to complex, high-dimensional data. Conversely, some methods are effective at identifying clusters with irregular shapes but require sensitive parameter tuning or exhibit higher computational costs. Therefore, the choice of clustering technique should be guided by the specific characteristics of the dataset, the objectives of the analysis, and evaluation criteria such as accuracy, computational efficiency, interpretability, and robustness to noise. Overall, this review aims to serve as a comprehensive reference for researchers, practitioners, and decision-makers in selecting the most suitable clustering method for their specific analytical needs. Additionally, the study highlights potential avenues for future research, including the development of hybrid algorithms, automated parameter selection techniques, and the integration of clustering with modern machine learning approaches to enhance performance and adaptability across various data environment
Developing a Digital Circular Economy Business Model for SIPETRA in Jatiluwih Village, Bali
No full textWaste management in tourism villages has become a major environmental challenge due to increasing waste generation from tourism activities and limited local infrastructure. Jatiluwih Village, a UNESCO World Heritage site in Bali, faces this issue as tourism growth produces more domestic and organic waste. Although several studies have examined community-based waste initiatives, research integrating strategic analysis, participatory validation, and business model innovation remains limited. This study aims to design a sustainable business model for SIPETRA (Waste Management and Technology System for Waste Transformation) as a community-based circular economy solution supported by digital technology. Unlike previous studies, this research integrates SWOT, Delphi, BMC, and BOS into a unified framework to develop a digital circular economy model tailored to rural tourism contexts. The research employed a descriptive qualitative method through field observation, interviews, questionnaires, and literature analysis. Data were processed using SWOT and Delphi techniques to identify strategic factors, followed by the formulation of the Business Model Canvas and Blue Ocean Strategy. The results show that SIPETRA’s internal capacity is moderately weak (IFAS = 2.45), while external opportunities are strong (EFAS = 3.01), placing the model in the WO quadrant. Consensus from 12 experts (Kendall’s W = 0.78) identified four strategic priorities: human resource improvement, digital transformation, product quality enhancement, and partnership-based funding. The BOS analysis generated innovative programs such as the SIPETRA app, Eco-Coin reward system, and Green Experience Center to create a “Jatiluwih Circular Living Experience.” This study concludes that the integrated analytical framework effectively transforms waste management into a self-sustaining digital circular economy model that supports environmental sustainability, social empowerment, and green tourism. The findings provide theoretical contributions to digital circular economy literature and practical implications for tourism villages seeking scalable and community-driven waste management solutions
Performance and Energy Efficiency Assessment of Embedded Arduino Vibrating Sieving System for Dry Powder Materials
No full textDry powder sieving is a crucial process for micro, small, and medium enterprises (MSMEs), where particle uniformity directly impacts product quality and production efficiency. Traditional vibrating sieving machines in local markets are typically evaluated through visual inspection, resulting in subjective assessments without quantitative evidence of energy efficiency or vibration stability. High humidity often causes powder clumping, reducing consistency and reliability. To address these limitations, this study introduces an Arduino based embedded system for quantitative performance and energy evaluation of a vibrating dry powder sieving process. System integrates an Atmega328P microcontroller (Arduino Uno) with infrared and DHT11 sensors, an L298N motor driver, a DC motor, and an LCD display. Electrical parameters (voltage and current) and vibration signals (acceleration along the X, Y, and Z axes) were acquired in real time at a sampling frequency of 10 Hz and recorded to an SD card for 60–90 seconds per cycle. Metrics included electrical power, energy consumption, vibration RMS, peak amplitude, dominant frequency, and energy efficiency expressed as the mass of powder sifted per joule of energy consumed. Experimental results, conducted using rice flour as a representative dry powder, showed that high humidity increased agglomeration, while a reciprocating motor at 210 RPM improved particle distribution across the sieve. The infrared sensor reduced energy consumption by activating the motor only when material was present. Overall, the system achieved an efficiency improvement exceeding 85% compared to manual sieving. This study demonstrates that embedded sensing and data acquisition can transform traditional sieving machines into objective, transparent, and reproducible systems for MSMEs, with potential application to various dry powder
The IT GOVERNANCE IN REGIONAL WATER COMPANY RISK MANAGEMENT USING THE COBIT 2019 METHOD
No full textDigital transformation in the public utility sector, particularly in regional water-owned enterprises (BUMD), presents complex risk challenges ranging from cybersecurity threats to operational distribution disruptions. PT Tirta Sriwijaya Maju (Perseroda), as the research object, faces constraints in IT risk management processes that are currently manual, reactive, and disintegrated, potentially threatening the sustainability of public services. This study aims to evaluate the current IT governance capability and design risk management improvements using the COBIT 2019 framework. The research methodology employs a mixed-method approach utilizing the Design Toolkit to determine domain priorities based on the company's risk profile and strategy. The evaluation focuses on six critical domains: EDM03, APO12, APO13, BAI03, DSS01, and MEA01. The Design Factors analysis established a target capability at Level 3 (Defined Process) to ensure regulatory compliance. However, the current state (As-Is) measurement indicates that the company is at an average of Level 1 (Performed). A gap of 2 levels was identified, primarily caused by a disconnected evaluation cycle (MEA01), the absence of a formal Risk Appetite document, and reliance on spreadsheet-based risk monitoring. As a solution, this study provides strategic recommendations including the formalization of risk policies, the design of an integrated digital Monitoring Dashboard, and an Implementation Roadmap for 2025-2027. The implementation of this roadmap is expected to enhance risk governance maturity, ensure customer data integrity, and guarantee operational stability in accordance with Good Corporate Governance standards.
Keywords: IT Governance, Risk Management, COBIT 2019, Design Factors, Regional Water Utility, Capability Level
Comparison Of Adam and SGD For The Classfication Of Palm Tree Leaf Diseases With ResNet50
No full textPlants from the palm tree family (Arecaceae), such as coconut, oil palm, and date palm, play an important role in the economy and food security, especially in Indonesia. However, leaf diseases such as leaf spot disease pose a serious threat that can reduce productivity. Manual disease identification is time-consuming and prone to errors, necessitating an image-based automatic classification system. This study aims to apply the ResNet50 Convolutional Neural Network (CNN) architecture for palm tree leaf disease classification and compare two popular optimization algorithms, Adam and Stochastic Gradient Descent (SGD), in terms of model training accuracy and efficiency. The dataset used is public, covering five classes of leaf images: Healthy, White Scale, Brown Spot, Leaf Smut, and Bacterial Leaf Blight. The research process includes data collection and preprocessing (resizing, normalization, and augmentation), dividing the dataset into three parts, namely training, validation, and testing data using the train/validation/test split approach. This approach provides a fairly representative evaluation of model performance while being computationally efficient. Model training was performed using transfer learning with ResNet50, and performance evaluation was performed using a confusion matrix to obtain accuracy, precision, recall, and F1-score values. The results of the two optimizers were compared to determine their effect on model performance. The experimental results show that the ResNet50 model optimized with Adam achieved a higher test accuracy of 87.23% compared to SGD with 85.96%, while SGD demonstrated more consistent performance between validation and testing phases, indicating better training stability
Deep Learning-Based Multi-Tooth Segmentation on Panoramic Radiographs Using YOLOv8 Architecture
No full textThis research introduces a multi-class tooth-level segmentation framework on panoramic radiographs using YOLOv8, trained on clinically annotated Indonesian dental data. A dataset of 302 annotated panoramic radiographs from patients at Universitas Andalas Dental Hospital was utilized, with each tooth precisely labeled according to international dental nomenclature. The model was trained using transfer learning with the YOLOv8 variant, optimized with the Adam algorithm, and evaluated using precision, recall, F1-score, and Intersection over Union (IoU). The results demonstrate that YOLOv8 is not only effective for lesion detection but also robust for fine-grained anatomical dental segmentation. The performance achieved 93.72% accuracy, 92.67% precision, 98.88% recall, and 95.58% F1-score, indicating high accuracy in tooth detection and boundary delineation. Qualitative analysis confirmed accurate segmentation across a wide range of anatomical variations, including crowding, impaction, and prosthetics. This research establishes YOLOv8 as a highly effective tool for dental image segmentation, offering significant potential to improve diagnostic efficiency, support odontological forensics, and enable automated patient record management. Future work will focus on integrating multi-class pathology detection and 3D reconstruction
An Optimized Lightweight CNN with Randomized Hyperparameter Search for Real-Time Image-Based Malware Detection
No full textWhile image-based malware detection using deep learning has shown promise, existing methodologies predominantly rely on computationally expensive pre-trained architectures (e.g., VGG, ResNet) that create significant bottlenecks for real-time deployment on resource-constrained gateways. This research addresses this critical gap by proposing a streamlined, lightweight custom Convolutional Neural Network (CNN) specifically optimized for real-time operation. The novelty of this work lies in the strategic integration of Randomized Search Cross-Validation (RS-CV) to automate the discovery of an optimal configuration of filters, dense units, and dropout rates, eliminating the inefficiencies and biases of manual hyperparameter tuning. The proposed method transforms binary files into 64x64 grayscale images—reducing computational input by over 90% compared to standard architectures—which are then processed by the optimized custom network. Experimental results demonstrate the scientific significance of this approach, as the model achieved a near-perfect Area Under the Curve (AUC) of 0.9996 and identified threats with an average inference time of only 12–15 milliseconds. Out of 1,068 test samples, only 10 misclassifications were recorded, proving that a mathematically optimized lightweight model can outperform heavy ensemble frameworks in both accuracy and speed. These findings provide a reproducible framework for high-speed, front-line cybersecurity systems capable of detecting obfuscated threats in live network environments
Generalized Chatterjea Type Contractions on Integrated Matrix Graph Metric Spaces
No full textThis paper proposes a computationally verifiable integrate fixed point framework on the integrated metric space , where combines a continuous component endowed with the matrix induced metric with invertible and a discrete component defined by the shortest-path metric of a finite weighted graph. The objective is to obtain verifiable conditions that guarantee existence, uniqueness, and predictable convergence of fixed points for coupled continuous–discrete dynamics, while embedding the graph geometry directly into the metric via the scaling parameter . Our method studies the coupled operator and derives explicit sufficient inequalities ensuring that satisfies a Chatterjea-type contraction on , yielding an effective contraction factor . In particular, the threshold implies that admits a unique fixed point and that the hybrid Picard iteration converges geometrically in . Numerical experiments support these findings and clarify the integrate mechanism, when maps every vertex to a fixed node, the discrete mode stabilizes after the first iterate, and the successive iterate error decays exponentially at a rate consistent with , with numerical and analytic fixed points agreeing up to floating-point tolerance. Practically, the bound provides an a priori, computable convergence for implementations of matrix graph iterations relevant to graph structured computing and networked models. Future work includes reducing conservatism in the sufficient bounds, exploring richer couplings, and extending the analysis to broader graph classes
The DESIGN FACTOR THE CAPABILITY LEVEL OF INFORMATION TECHNOLOGY GOVERNANCE AND RECOMMENDATIONS FOR IMPROVEMENT USING COBIT 2019: A CASE STUDY AT A PRIVATE UNIVERSITY IN PALEMBANG
No full textPrivate university in Palembang that has utilized information technology to support its educational processes. This study demonstrates how COBIT 2019 design factors can be operationalized to identify strategically critical governance domains in higher education, shifting IT governance evaluation from generic domain assessment toward contextualized governance design, This study designs an information technology governance system based on COBIT 2019 and applies design factors to align governance objectives with university institutional context and strategic direction, thereby identifying priority domains and capability level targets that are most relevant to the university’s digital transformation. The application of these design factors highlights the EDM05 Ensure Stakeholder Engagement and BAI09 Manage Assets domains as the primary focus, as they are closely related to the need to enhance transparency in stakeholder involvement and to optimize IT asset management within the university environment. The study aims to assess the capability level of IT governance in the EDM05 and BAI09 domains, identify gaps between the current conditions and the targeted capability levels, and formulate improvement recommendations that are aligned with institutional needs and strategies. The research adopts a case study approach with a descriptive mixed method design, employing observation, interviews, and questionnaires, which are then analyzed using RACI Chart mapping, COBIT 2019 capability level measurement, and gap analysis to develop proposed process improvements. The findings indicate that the capability levels in both domains are still below the targeted levels, resulting in gaps related to stakeholder engagement, role and responsibility structures, IT performance reporting mechanisms, and IT asset lifecycle management. The implications of this research are the provision of a practical foundation for university in Palembang’s management to strengthen IT governance, as well as an academic contribution on the application of COBIT 2019 design factor–based governance in higher education institutions that can serve as a reference for other universities.
Keywords : IT Governance, COBIT 2019, Capability, Recommendatio
Designing and Evaluating a User-Centered Cash Flow Monitoring Dashboard for Higher Education Using Design Thinking and UEQ Framework
No full textAlthough cash flow monitoring dashboards have been widely implemented in higher education institutions, existing studies predominantly focus on technical development or usability testing, with limited attention to how user-centered design frameworks contribute to financial decision-making effectiveness. This creates a research gap regarding the systematic application of design thinking as a methodological approach for developing and evaluating financial dashboards in university contexts. This study addresses this gap by proposing and evaluating a user-centered cash flow monitoring dashboard developed using a design thinking approach. A descriptive quantitative method was employed through a case study conducted at Sekolah Tinggi Teknologi Terpadu Nurul Fikri. Data were collected through stakeholder interviews and prototype evaluation using the User Experience Questionnaire (UEQ). The design thinking process was implemented across the empathize, define, ideate, prototype, and test stages to ensure alignment between user needs and dashboard functionality. The findings indicate that the proposed dashboard achieved strong performance across key UEQ dimensions, particularly attractiveness, efficiency, and dependability, demonstrating its effectiveness in supporting cash flow monitoring and managerial financial decision-making. Unlike previous studies that emphasize system implementation outcomes, this research provides empirical insights into how design thinking facilitates the translation of user needs into actionable financial information. This study contributes to the literature by offering a structured framework for applying design thinking in the development of financial monitoring dashboards within higher education institutions. The results also have practical implications for universities seeking to improve data-driven financial governance through user-centered financial information systems