UMA - Open Access Journals (Universitas Medan Area)
Not a member yet
5228 research outputs found
Sort by
Hybrid CNN-LSTM for Indonesian Cyberbullying Detection on Social Media X
Full text linkCyberbullying on social media platform X has become a critical digital threat and requires automatic detection mechanisms to mitigate psychological impacts on victims. This study proposes a hybrid deep learning architecture that combines Convolutional Neural Network (CNN) for local feature extraction and Long Short-Term Memory (LSTM) for sequential context understanding in classifying Indonesian language cyberbullying comments. This study evaluates model performance using a dataset of 13,677 comments from social media X through a series of systematic testing scenarios, including the impact of regularization, utilization of FastText embeddings, and comparative studies against state-of-the-art models. Experimental results demonstrate that the Early Stopping mechanism is a critical factor in this architecture, where without this mechanism the model experiences accuracy degradation of up to 32%. The proposed CNN-LSTM model achieves 88.38% accuracy and 88.00% F1-Score, improving to 0.9559 AUC with FastText integration. This model achieves over 97% of IndoBERTweet's performance with 22 times lower computational complexity (4.97 million versus 110.88 million parameters) and outperforms machine learning methods such as SVM with an accuracy margin of more than 10 percentage points. This study concludes that the CNN-LSTM architecture offers a robust and efficient solution for cyberbullying detection, particularly for resource-constrained environment
The Comparison Of Decision Tree And Naive Bayes Algorithms In Classifying Heart Disease
No full textAbstract
Corony Heart Diease (CHD) is one of the most common causes of death worldwide, and early detection is crucial to reduce its fatal effects, In this study, two Machine Learning algorithms, Decision Tree and Naïve Bayes, were evaluated to classify heart disease using the Kaggle dataset. These algorithms were evaluated based on their accuracy, precision, recall, and fi-score across three training and testing data proportions (85:15, 75:25, and 60:40). The result showed that Decision Tree consistenly outperformed Naïve Bayes. The superiority of Decision Tree was also demonstrated by metrics such as precision dan recall. The main advantage of Decision Tree is its ability to handle both numerical and categorical data and make classification results easier to understand. In contracts, Naïve Bayes showed limitations in indentifying complex patterns in medical data. Despite its success, the evaluation results showed that the Decision Tree algorithm is better for heart disease classification because it is more stable and accurate. This study provides insights for developers of more efficient early detection systems for heart disease using Machine Learning
NANOEMULSION-BASED BIOPESTICIDE FROM LEMONGRASS, GARLIC, AND CIGARETTE BUTTS FOR CONTROLLING Fusarium WILT IN PATCHOULI (Pogostemon Cablin)
Full text linkDiseases of patchouli (Pogostemon cablin) caused by fungal pathogens are a major constraint to crop productivity and oil quality, highlighting the need for effective and environmentally friendly control strategies. This study aimed to evaluate the physicochemical characteristics, antifungal activity, and in vivo effectiveness of fermented biopesticide formulations derived from garlic (Allium sativum), lemongrass (Cymbopogon citratus), and cigarette butts. The research consisted of formulation and fermentation using EM4, analysis of physicochemical properties, in vitro antifungal assays against two fungal pathogens at different concentrations, and in vivo evaluation of disease incidence and plant growth. The results showed that fermentation significantly reduced pH, indicating active microbial metabolism. In vitro tests demonstrated that both formulation type and concentration significantly affected mycelial growth inhibition, with formulations F1 and F2 exhibiting the highest antifungal activity and near-complete inhibition at higher concentrations. In vivo application confirmed these results, as F1 and F2 significantly reduced disease incidence and improved plant growth compared to the control. In conclusion, fermented biopesticide formulations, particularly F1 and F2, have strong potential as effective and environmentally friendly alternatives for managing patchouli diseases and supporting sustainable agriculture
Implementation of the Responsibility to Protect Principle in the Prevention of Genocide in Indonesia
Full text linkThis study aims to analyze the implementation of the Responsibility to Protect (R2P) principle in the context of genocide prevention. Using a qualitative method based on literature study, this study examines national legal documents such as the Human Rights Law, the Social Conflict Handling Law, and Indonesia's diplomatic statements in UN forums related to R2P. Data were analyzed using Miles and Huberman's interactive models to identify the fit between R2P standards and Indonesian policy instruments. The results show that Indonesia supports R2P normatively, especially in Pillar 1 which emphasizes the state's responsibility to protect citizens from mass violence. However, this support is ambivalent because the state still rejects Pillar 3 which opens up the possibility of international collective intervention. A national policy analysis revealed that Indonesian legal instruments have contained elements of protection, but are not specifically designed to prevent genocide, so their implementation is more reactive than preventive. The Papuan case study shows that there is a risk of structural genocide that has not been officially recognized by the state. The main obstacles to the implementation of R2P in Indonesia lie in the dominant security paradigm, the absence of a human rights-based early warning system, and the politics of denial of the potential for mass violence. This study recommends the explicit integration of R2P principles into national policies, the establishment of early warning systems, and the reorientation of security approaches to human security. In conclusion, Indonesia is at the stage of R2P as a discourse, but it has not fully become R2P as a policy practice
Empathic Communication and Behavioral Change in Stunting Prevention by Family Assistance Teams
Full text linkThis study examines how empathic interpersonal communication practiced by Family Assistance Teams (TPK) contributes to stunting prevention in Labuhanbatu Utara District, Indonesia. Using a qualitative case study design grounded in Interpersonal Communication Theory and the Communication for Development (C4D) approach, data were collected through in-depth interviews, participatory observation, and focus group discussions with TPK cadres, families at risk of stunting, and local policymakers. The findings demonstrate that empathic communication characterized by openness, responsiveness, and emotional sensitivity plays a critical role in fostering parental behavioral change related to child nutrition and health. Home visits, interpersonal dialogue, and the use of everyday language enhance families’ understanding and acceptance of healthy practices by strengthening trust and reducing resistance. However, the effectiveness of these practices is constrained by geographical barriers, social stigma, and limited communication training among cadres. This study’s main contribution lies in highlighting empathic interpersonal communication as a key behavioral-change mechanism within community-based stunting prevention, reinforcing C4D as a practical framework for advancing Indonesia’s zero-stunting agenda
Smart Mobility : The Government policy to boost the tourism sector through the Heritage Track Bus in Yogyakarta City
Full text linkThis article examines the implementation of smart mobility policies through Heritage Bus Routes as a government strategy to strengthen sustainable tourism development in the city of Yogyakarta in the context of Society 5.0. This study analyzes how transportation innovations can be integrated with cultural heritage preservation to increase tourist participation and enhance visitor experiences in heritage cities. Using a descriptive qualitative approach, this study analyzes policy implementation, stakeholder collaboration, and user perceptions. Data were collected through in-depth interviews with key stakeholders, program users, and educators, and were supplemented by secondary data from the literature and official documents. This study was conducted between August and November at the Yogyakarta Special Region Cultural Office. Findings indicate that the heritage bus policy has largely achieved its objectives, reflected in tourist participation exceeding targets, route expansion, and consistent user satisfaction. Despite challenges such as limited fleet availability and a suboptimal reservation system, the program effectively integrates urban transportation with cultural heritage preservation along the Yogyakarta Philosophy Axis through educational tourism. Supported by strong stakeholder collaboration and positive public perception, the Heritage Bus represents an innovative model of smart mobility and provides a policy reference for other heritage cities.Transformasi era society 5.0 menuntut inovasi pengelolaan pariwisata berkelanjutan melalui penerapan smart mobility. Penelitian ini bertujuan untuk menganalisis implementasi Smart Mobility melalui program Bus Heritage Track sebagai kebijakan pemerintah dalam peningkatan sektor pariwisata di Kota Yogyakarta. Metode penelitian yang digunakan adalah kualitatif deskriptif dengan pengumpulan data primer melalui wawancara terhadap stakeholder, pengguna, dan edukator Bus Heritage Track, serta data sekunder dari studi literature dan kajian Pustaka. Penelitian dilaksanakan pada bulan Agustus-November di Dinas Kebudayaan Daerah Istimewa Yogyakarta. Hasil penelitian menunjukkan bahwa implementasi kebijakan Bus Heritage Track berhasil mencapai target dengan peningkatan jumlah wisatawawan yang melebihi ekspentasi awal, efektivitas program dengan penambahan jumlah track setiap tahunnya dan tingkat kepuasan pengguna yang menunjukkan nilai positif pada tahun 2021 hingga 2025. Penelitian ini menyimpulkan bahwa Bus Heritage Track merupakan program inovatif yang berhasil mengintegrasikan moda transportasi dengan pelestarian warisan budaya, terkhususnya Sumbu Filosofi Yogyakarta melalui konsep wisata edukatif. Meskipun menghadapi kendala dari minimnya reservasi dan keterbatasan armada. Keberhasilan program ini ditentukan oleh kolaborasi antar stakeholder, respon positif pengguna dan kemampuan memberikan pengalaman wisata heritage yang bermakna bagi pengguna, sehingga menjadi model strategis dalam pengembangan pariwisata berkelanjutan di Kota Yogyakarta
GILLS VISUALIZATION OF TILAPIA AND MORTALITY RATES USING THE LETHAL 50 METHOD TO DETECT SUBLETHAL OIL POLLUTION IN RIVER WATER BODIES
Full text linkEnvironmental issues in Indonesia are mostly caused from pollution resulting from industry and domestic waste. Oil and used cooking oil are examples of waste that are often produced from both. Water contaminated by these chemicals can affect organisms living in it, one of which is tilapia (Oreochromis niloticus). Hazardous chemicals can accumulate in tilapia through the food chain, absorption through the gills, or diffusion through the skin surface, potentially causing death. This study aims to determine the effect of used cooking oil and motor oil waste on gill morphology and mortality of tilapia. The study was conducted for 96 hours using an experimental method by testing the effect of used cooking oil and used oil with concentrations of 1%, 3%, 5% on tilapia (O. niloticus). The parameters observed in the study were gill morphology and calculating the number of fish mortalities. The results show that increased concentrations of used cooking oil and motor oil increased tilapia mortality, with the highest mortality at a concentration of 5%, indicating that exposure had exceeded the sublethal threshold and was lethal
ENVIRONMENTAL MANAGEMENT PERFORMANCE AND GOVERNANCE CHALLENGES IN ROCK MINING OPERATIONS IN BONE BOLANGO REGENCY
Full text linkRock mining plays an important role in supporting regional infrastructure development in Bone Bolango Regency, Gorontalo Province. However, quarry expansion also poses environmental risks that require proper management and supervision. This study evaluates the implementation of environmental management at three active rock mining sites. A mixed-methods approach was applied, integrating field observations, interviews, document analysis, and ambient air quality monitoring to assess compliance with UKL-UPL requirements. The results indicate that environmental management implementation remains inadequate. Major issues include weak erosion and sediment control, limited dust suppression, insufficient waste management, and inconsistent environmental monitoring, despite the availability of formal management documents. Air quality measurements at AF-01, AF-02, and AF-03 show spatial variation in particulate concentrations, influenced by mining intensity and local climatic conditions. Although all values are below national ambient air quality standards, field observations reveal localized dust accumulation, indicating insufficient on-site mitigation. Governance challenges further constrain the application of Good Mining Practices, including limited enforcement capacity, the absence of certified KTT/PJO personnel, and weak inter-institutional coordination. Local communities also reported disturbances related to dust, noise, and truck traffic. Overall, the study emphasizes the need for stronger regulatory oversight, improved technical capacity, and participatory monitoring to support adaptive and sustainable mining management
Integrating Automatic Stock Monitoring and Digital Inventory Systems for MSMEs A Mobile Application Approach (Case Study in Serang City, Indonesia)
Full text linkMicro, Small, and Medium Enterprises (MSMEs) in Indonesia continue to face inefficiencies in inventory management due to manual stock recording, data inconsistency, and delays in operational decision-making. In Serang City, these challenges often lead to stockouts, excess inventory, and limited business scalability. This study aims to develop and evaluate a mobile-based automated inventory management system that supports real-time stock monitoring and decision-making for MSMEs. The research employs a Research and Development (R&D) approach integrated with the Agile-Scrum methodology, encompassing problem identification, user requirement analysis, system design, prototype development, functional testing, and usability evaluation. Functional validation was conducted using black box testing, while system usability was assessed using the System Usability Scale (SUS) involving 15 MSME users. The results indicate that all core system functions achieved a 100% success rate, including automated stock recording, cloud-based data synchronization, real-time notifications, and dashboard analytics. The usability evaluation produced an average SUS score of 82.5, classified as Excellent, indicating high user acceptance and ease of use. These findings demonstrate that the proposed system effectively improves inventory accuracy, operational efficiency, and decision-making quality, contributing to MSME digital transformation in developing regions
Improving Imbalanced Polycystic Ovary Syndrome Classification Using a Leakage-Free Machine Learning Pipeline
Full text linkPolycystic Ovarian Syndrome (PCOS) is a complex endocrine disorder affecting women of reproductive age and poses challenges for early diagnosis due to heterogeneous clinical presentations and imbalanced clinical datasets. This study aims to develop a data leakage–free machine learning pipeline to enhance the accuracy and reliability of PCOS classification using clinical data. The dataset underwent preprocessing and normalization, followed by stratified data splitting with an 80:20 ratio to maintain class proportions. The proposed pipeline was implemented within a unified computational framework integrating feature selection based on the ANOVA F-test, class imbalance handling using the Synthetic Minority Over-sampling Technique (SMOTE), and classification using a Support Vector Machine (SVM) with a Radial Basis Function (RBF) kernel. Hyperparameter tuning was performed using GridSearchCV combined with K-Fold Cross-Validation to ensure model robustness and consistency. The experimental results indicate that the proposed model achieved an accuracy of 0.9074, with precision, recall, and F1-score values of 0.8378, 0.8857, and 0.8611, respectively. Furthermore, ten dominant clinical features were identified, primarily related to hormonal profiles and ovarian morphology. These results demonstrate that the data leakage–free pipeline improves the validity and stability of PCOS prediction. The findings suggest that this approach may serve as a supportive tool for clinical decision-making, particularly in facilitating early and objective identification of PCOS