JKTI Jurnal Keilmuan Teknologi Informasi

Integration of 8 Golden Rules in Payroll Information System Interface Design

Hartika Rahayu — 2025-07-14

The development of a system is done to help facilitate human work due to its effectiveness and efficiency. The XZY campus study program has used an integrated system in processing data. In terms of summarizing attendance and payroll data, it is still done manually, so a system is needed that can help users summarize this data. The interface design construction is made before system development. There are three previous steps in designing the system display: data collection (needs analysis), website design, and testing. The 8 golden rules are used as guidelines in designing a web display that is easy for users to learn.

Deep Learning Approach for Sleepiness Detection with YOLOv12

Diesti Hidayani — 2025-07-14

Drowsiness while driving is a significant contributor to traffic accidents. To mitigate such occurrences, a precise and real-time drowsiness detection system is essential. This research aims to create a computer vision-based drowsiness detection system utilizing the YOLOv12 algorithm. The dataset was sourced from Kaggle and manually annotated with the help of Roboflow. It was categorized into two groups: drowsy and non-drowsy, with the original 5,000 images augmented to a total of 6,976 images. The model training utilized the AdamW optimizer (learning rate=0.001667, momentum=0.9) over 100 epochs and a batch size of 4. Performance assessment indicates that the model attained an mAP@50 of 0.732 and an mAP@50-95 of 0.62, alongside a precision of 0.648 and a recall of 0.928. These findings illustrate that YOLOv12 can successfully identify drowsiness in real-time. Nevertheless, the performance of the model is significantly influenced by the quality and balance of the dataset. Consequently, enhancing the structure and distribution of the dataset is vital for improving detection accuracy.

Implementation of Tesseract-based OCR for UMKLA student card data extraction

Muhammad Nashiruddin — 2025-07-14

Manual data entry from student ID cards (KTM) is often inefficient and prone to errors. Therefore, automating this process is a crucial solution for educational institutions to improve accuracy and the speed of administrative services. This research aims to design and implement an Optical Character Recognition (OCR) system to automatically extract information from student ID card images of Universitas Muhammadiyah Klaten (UMKLA). The methodology involves image pre-processing using the OpenCV library to enhance image quality through grayscale conversion and Otsu's binarization. Subsequently, the Tesseract OCR Engine is used to convert the image into raw text, which is then parsed using Regular Expressions (Regex) to separate data fields such as Name, Student ID Number (NIM), and Program of Study. Test results indicate that the system can extract information with a good success rate, although accuracy is heavily influenced by image quality factors like lighting and text clarity. Fields with standard printed formats were found to have higher accuracy. In conclusion, this Tesseract-based system successfully demonstrates its feasibility for local automation of student ID card data. However, further development in the post-processing stage is required to handle more complex OCR output variations.

Analytical Prediction for Chronic Kidney Disease: A Comparison of Machine Learning Methods

Krisna Nuresa Qodri — 2025-07-15

Chronic kidney disease (CKD) is a progressive malady defined by reduced glomerular filtration rate, increased urinary albumin excretion or both, and is a major global public health concern with an extremely high unmet medical need. CKD is estimated to occur in 8-16% of the worldwide population and results in a substantially reduced life expectancy. Early detection and accurate prediction of CKD is crucial to reduce health complications such as hypertension, anemia, and premature death. This study aims to develop CKD prediction models using three machine learning methods: Random Forest, Naive Bayes, and Support Vector Machine, then compare the performance of each method. The dataset used is the CKD dataset from UCI Machine Learning Repository consisting of 400 instances with 24 attributes. Experimental results show that Random Forest achieved 90.50% accuracy, Naive Bayes achieved the highest accuracy of 94.21%, while SVM achieved 88.84% accuracy. The results indicate that Naive Bayes provides the best performance for chronic kidney disease prediction with superior accuracy compared to other methods. This prediction model can assist medical practitioners in early detection and appropriate clinical decision-making for CKD patient management.

Real-Time Facial Expression Detection Using YOLOv12

Rizal Adimas — 2025-07-15

This research focuses on the development of a real-time facial expression detection system using YOLOv12. The study utilizes a secondary dataset from Kaggle, consisting of 1000 images categorized into two classes: "Happy" and "Not Happy." The dataset undergoes preprocessing steps, including Gabor filter bank for key facial feature identification and geometric augmentation to enhance data quality. The YOLOv12 model is trained with 100 epochs, a batch size of 4, and the AdamW optimizer, achieving a mean Average Precision (mAP@0.5) of 0.89 for both expression classes. The system demonstrates real-time performance with an average processing speed of 15 FPS on CPU-based devices, adapting well to varying lighting conditions and angles, though accuracy decreases by 5-7% in low-light environments. The results highlight the model's potential applications in mental health, human-computer interaction, and security. Limitations include the restricted dataset and challenges with micro-expressions. Future work suggests expanding the dataset to include more expression classes and integrating post-processing models to reduce false positives.