Sinkron : jurnal dan penelitian teknik informatika
Not a member yet
1261 research outputs found
Sort by
Fairer Public Complaint Classification on LaporGub: Integrating XLM-RoBERTa with Focal Loss for Imbalance Data
No full textThe advancement of digital technology has provided opportunities for governments to improve the quality of public services through citizen complaint channels. One example of this implementation in Indonesia is Lapor Gub, managed by the Dinas Komunikasi dan Informasi Provinsi Jawa Tengah (Communication and Information Agency of Central Java Province). This platform receives thousands of complaints daily, ranging from infrastructure, social issues, to illegal levies. However, the large volume of data and the imbalanced distribution of categories pose significant challenges for both manual and automated processing. This study aims to classify citizen complaint texts using XLM-RoBERTa combined with Focal Loss as an approach to handle data imbalance. The dataset consists of 53,774 complaints after data cleaning and text preprocessing. The training process applied a stratified split (78% training, 18% validation, 10% testing) and fine-tuning for 10 epochs. Model performance was evaluated using accuracy, precision, recall, and macro F1-score. The results show that the model without Focal Loss achieved 78.1% accuracy with a macro F1-score of 0.606, while the model with Focal Loss improved the macro F1-score to 0.625 with 78.5% accuracy. These findings demonstrate that the application of Focal Loss enhances the model’s ability to recognize minority categories without reducing performance on majority classes. Therefore, the combination of RoBERTa and Focal Loss offers an effective solution to support faster, fairer, and more transparent public complaint management
Designing a Stunting Prediction Model Using Machine Learning to Support SDGs Achievement in Indonesia
No full textStunting remains a major public health challenge in Indonesia, with national prevalence among children under five reaching 21.6% in 2022, according to the Ministry of Health. This condition, defined by the World Health Organization as a height-for-age less than -2 SD, is associated with long-term consequences including impaired cognitive development, reduced educational attainment, and diminished economic productivity. Addressing stunting is therefore critical to achieving Sustainable Development Goals (SDGs) related to hunger, health, and education. Despite multiple national initiatives, early identification of stunting risk is still limited by reliance on conventional, reactive surveillance methods. Recent advances in machine learning (ML) provide promising alternatives for proactive stunting prediction, with several studies reporting high predictive accuracy using ensemble methods, hybrid frameworks, and geographically weighted models. Building upon this evidence, the present study develops and evaluates ML models for stunting risk prediction using a large dataset of 10,000 records from North Sumatra, Indonesia. The dataset included three predictor variables—age, height, and weight—and a target variable, nutritional status (Normal, Stunted, Severely Stunted, Tall). Four algorithms were compared: K-Nearest Neighbors (KNN), Naïve Bayes, Decision Tree, and Random Forest. Performance was assessed using accuracy, precision, recall, F1-score, and ROC area, with 10-fold cross-validation ensuring robust estimation. Results demonstrated that Decision Tree (88.6% accuracy) and Random Forest (88.3% accuracy) outperformed KNN (84.7%) and Naïve Bayes (72%). ROC areas further confirmed the superiority of ensemble-based approaches, particularly Random Forest (0.979). Statistical significance was tested using McNemar’s test, revealing that Decision Tree and Random Forest achieved comparable performance (p = 0.651), both significantly outperforming KNN and Naïve Bayes (p < 0.05). This study contributes a context-specific evaluation of ML methods for stunting prediction in North Sumatra, emphasizing not only predictive accuracy but also interpretability to support health policy and program implementation. By bridging data-driven insights with actionable decision support, the proposed framework advances progress toward SDG-aligned strategies and provides a foundation for more targeted and preventive interventions in child nutrition and growth monitoring
Enhanced Semarang Batik Classification using MobileNetV2 and Data Augmentation
No full textBatik, an Indonesian cultural heritage recognized by UNESCO, faces challenges in pattern identification and documentation, particularly for the younger generation. Previous studies on batik classification have shown limitations in handling small datasets and maintaining accuracy with limited computational resources. This research proposes an enhanced classification approach for Semarang Batik motifs using MobileNetV2 architecture combined with strategic data augmentation techniques. The study utilizes a dataset of 3,020 images comprising 10 distinct Semarang Batik motifs, implementing horizontal flipping, rotation, and zoom transformations to address dataset limitations. Our methodology incorporates transfer learning through ImageNet pre-trained weights and custom layer modifications to optimize the MobileNetV2 architecture for batik-specific features. The model achieves 100% accuracy on validation data, with precision, recall, and F1-scores consistently above 0.98 across all classes. The confusion matrix analysis reveals minimal misclassification between similar motif patterns, particularly in the Batik Blekok Warak and Batik Kembang Sepatu classes. This research contributes to cultural heritage preservation by providing an efficient, resource-conscious solution for automated batik pattern recognition, potentially supporting educational and commercial applications in the batik industry
Comparison of C4.5 & Random Forest Based on AdaBoost For Determining Loan Eligibility Customer Funds
No full textThis research discusses the comparison between two data mining algorithms, namely Decision Tree C4.5 and Random Forest based on AdaBoost, in determining the creditworthiness of customer funds. The main objective of this research is to evaluate and compare the performance of the two algorithms in predicting loan eligibility based on customer data. Algorithm performance is measured using accuracy, precision, recall, and misclassification error metrics. The research results show that the AdaBoost-based Random Forest is superior with an accuracy of 78.86%, recall of 98.75%, and the lowest misclassification error of 21.14%. Meanwhile, Decision Tree C4.5 provides lower performance than AdaBoost-based Random Forest. This research recommends further exploration of other algorithms, such as Support Vector Machine (SVM) and Neural Networks, to obtain more optimal results in determining customer loan eligibility
Research on Sobel Edge Detection Algorithm of Grayscale Images to Analyse Car Number Plate
No full textImage processing is a very important subject to be discussed in computer science. Many applications of image processing are already in the field. Image processing techniques are applied in color and grayscale images. The application of image processing are ranging for military, medical and many other applications. One most important thing to analyse image and enhance its quality is doing edge detection. Edge detection in image is a well known approach to be used to detect discontinuity in grayscale image. Edge detection functions to identify edge line in images. Sobel algorithm is one of most known algorithm, others are prewitt, canny, homogeneity algorithms. Image can be made sharper and will enhance its quality. To detect number plate of cars, an edge detection algorithm needs to be applied. In number plate, to recognize the cars number plate, the image should be clear and clean from dirt. Sometimes we can not recognize the plate number if it is too blur or has many dirt. So in its application we need a strong edge detection algorithm to recognize car number plate easily. In this journal, five car’s images are presented. Each with the original image, grayscale image and the image after edge detected by sobel algorithm. It is concluded that this algorithm is quiet good in the implementation. But in the result, there are poor quality image also. For PSNR of images after edge detected, their values are between 19 and 20 dB, which are not good
Application of Extreme Programming Methods in the Design and Building of the Nusantara Capital Sentiment Analysis System
No full textInformation about the capital city of the archipelago (IKN) in the digital era serves as a platform for individuals to express views on development, policies, and socio-economic impacts. Such information often contains personal emotional expressions, categorized as negative, neutral, or positive sentiments. This study aims to design a sentiment analysis system to evaluate public opinions regarding IKN. The system utilizes Google NLP services, which offer sentiment measurement features for analyzed text, and web scraping techniques to automate data collection from online sources. The development process employs the Laravel framework and follows the Extreme Programming approach, which ensures work efficiency. Sentiment analysis is conducted using the Support Vector Machine (SVM) method, achieving an accuracy rate of 95%. The system is designed to be web-based, ensuring accessibility across devices, including smartphones and computers. The results demonstrate that this sentiment analysis system can help individuals, organizations, and governments gain deeper insights into public perspectives on IKN. Furthermore, it serves as a valuable tool for strategic decision-making and policy evaluation related to IKN development. Future research may explore expanding the data sources and integrating more advanced analytical techniques to improve system performance
Data Mining Framework for EDC Terminal Repair Protocol: Combining Apriori and PrefixSpan
No full textElectronic Data Capture (EDC) terminals are vital for financial transactions, but their repair processes often lack standardization, causing inefficiencies. Data mining techniques like Association Rule Mining (ARM) and Sequential Pattern Mining (SPM) can extract hidden patterns from service logs to inform maintenance strategies. This research addresses the limited use of these techniques within Electronic Data Capture (EDC) repair centers. Specifically, it applies Association Rule Mining (ARM) using the Apriori algorithm, and Sequential Pattern Mining (SPM) using the PrefixSpan algorithm, to optimize repair protocols based on historical repair data from PT. XYZ Indonesia. The study aimed to discover frequent fault-action-component associations and repair sequences to formulate standardized procedures. A quantitative case study analyzed 56,629 repair transactions. After data cleaning and transformation, Apriori (evaluated by support, confidence, lift) mined association rules, while PrefixSpan found frequent sequential patterns (evaluated by minimum support). Several high-confidence rules emerged: "Battery Not Charging" almost always led to "Replace Battery Pack" (≈95% confidence, lift ≈6.0), and error "2000000" (tamper indication) strongly correlated with detampering procedures and internal battery replacement (≈96% confidence, lift ≈4.9). PrefixSpan uncovered consistent repair sequences, including length-3 patterns for complex issues, with "Replace CMOS → Reinstall OS" for error "7FFFFF" being a prominent shorter sequence. Integrating these data-driven patterns into protocols and aligning inventory can improve service efficiency, reduce repair time, and enhance EDC reliability
Customer Loyalty Classification Using KNN and Decision Tree for Sales Strategy Development
No full textCustomer loyalty is a crucial element in maintaining business continuity in today’s competitive digital era. This study aims to classify customer loyalty levels based on sales and transaction behavior data using two supervised machine learning algorithms: K-Nearest Neighbor (KNN) and Decision Tree. The models were developed and evaluated using Python in the Google Colaboratory environment, utilizing a dataset of 250 customer records. The research process included data preprocessing, feature selection, normalization, data splitting, model building, and evaluation using accuracy, precision, recall, and F1-score metrics. Evaluation results showed that the Decision Tree algorithm delivered the best performance with 99.20% accuracy, 99.50% precision, 99.50% recall, and a 99.50% F1-score. Meanwhile, the KNN algorithm achieved 91.60% accuracy, 91.63% precision, 98.50% recall, and a 94.91% F1-score. These findings indicate that the Decision Tree model is more effective for classifying customer loyalty and can be implemented as a decision support tool for data-driven Customer Relationship Management (CRM) strategies
Optimizing Supplier Selection Through Hybrid BWM and AHP Integration
No full textThis study proposes a hybrid decision-making model that integrates the Best-Worst Method (BWM) with the Analytic Hierarchy Process (AHP) to optimize supplier selection. The primary objective is to address limitations in traditional Multi-Criteria Decision-Making (MCDM) methods, such as inconsistency, subjectivity, and cognitive overload when handling complex criteria. The proposed model leverages AHP's hierarchical structuring and BWM’s efficiency in reducing comparison load, aiming for a more accurate and consistent evaluation framework. The research design involves developing a hybrid AHP-BWM model and applying it to a dataset from the Vietnamese Textile and Apparel (T&A) sector. The methodology includes two stages: determining the weight of each criterion using a Hesitant-AHP approach, followed by evaluating supplier alternatives with BWM. The performance of the model is assessed using classification metrics, namely accuracy, precision, recall, and F1-score. The results show that the proposed model outperforms conventional methods such as TOPSIS, ELECTRE, VIKOR, and SWARA. It achieves an accuracy of 92%, precision of 87%, recall of 86%, and an F1-score of 86%. These outcomes confirm the model’s superior ability to consistently classify supplier suitability. Furthermore, the model identifies Quality Assurance as the most critical criterion, followed by Assistance, Capacity, Charge, and Shipment. In conclusion, the hybrid AHP-BWM model offers a robust, scalable, and data-driven approach for supplier selection. Its strength lies in balancing systematic evaluation with reduced cognitive effort, making it suitable for complex real-world decision-making environments. Future research may explore its application in other domains and enhance its scalability for larger datasets
Sentiment Analysis of Roblox Game Reviews Using Support Vector Machine Method
No full textThe development of digital technology has driven changes in entertainment consumption patterns, especially among the younger generation. Roblox has become one of the most popular online gaming platforms, with a wide range of user opinions recorded on Google Play Store. This study aims to classify the sentiment of Roblox user reviews (positive, negative, neutral) and evaluate the performance of the Support Vector Machine (SVM) algorithm with TF-IDF weighting and automatic labeling using Lexicon InSet. Data was obtained by crawling 10,000 reviews during the period of April 2–May 23, 2025, and after the preprocessing stage, 8,950 data remained for analysis. The classification results show that the sentiment distribution consists of 41.3% positive (3,703 reviews), 41.8% neutral (3,739 reviews), and 16.8% negative (1,507 reviews). Model evaluation using a confusion matrix produced high performance with 87.03% accuracy, 87.29% precision, 87.03% recall, and an F1-score of 86.67%. WordCloud visualization shows that positive reviews emphasize creativity and interactive features, while negative reviews are dominated by technical complaints such as lag and errors. These findings prove that the combination of SVM, TF-IDF, and Lexicon InSet is effective in sentiment analysis and provides valuable input for developers to improve application quality and user protection. Further research is recommended to adopt a hybrid approach based on deep learning and aspect-based sentiment analysis to generate more insights