Journal of Applied Computer Science and Technology

Pengembangan Aplikasi Augmented Reality untuk Edukasi Keselamatan Kebakaran: Metode Prototyping dan Usability

Ridwan Setiawan, Sri Rahayu, Iis Komalasari — Sun, 22 Jun 2025 14:27:06 +0100

This study aims to develop an Augmented Reality (AR) application to introduce fire safety equipment as part of Occupational Health and Safety (OHS) education for the general public. The research employed a prototyping method, which involved iterative stages from requirements analysis to user evaluation, as well as functional testing using Black Box Testing and usability testing with 20 respondents. The results showed that all application features functioned according to specifications, and the usability testing yielded a user satisfaction score of 80.1%. These findings indicate that AR is effective as an interactive educational medium to enhance public understanding of fire protection equipment. The implication of this study is the potential for wider use of AR technology in public education to support fire risk mitigation efforts.

Konsistensi Model Regresi Empat Variabel Pada Populasi dan Sampel untuk Prediksi Temperatur

Nurjannah Syakrani, Naufal Athaya S. R — Sun, 22 Jun 2025 00:00:00 +0100

The ability to predict future events or trends has become very important today. One method that can be used to predict the future is to use linear regression. Accurate regression modeling requires sampling representative data, especially when working with large datasets. This research takes a relatively large volume as a data set by looking at the accuracy and consistency of the coefficients of a multi-variable linear regression model for temperature prediction which is built based on all the data, and looks at the differences in the regression model built from the sample data. The number of sample data (n) is determined based on the Slovin formula which depends on the number of population data (N) and the level of confidence (ơ), so that as many as (N/n) new regression models can be built. Each group of sample data is divided into 75% for modeling and 25% testing data. The dataset used is weather information in the Szeged area which was measured in 2006 - 2016. So the regression model is in the form of Y (temperature value) which is influenced by Xi (weather factors), namely humidity, wind speed, wind direction and visibility. Using 96,453 data records and a 1% significance level in Slovin's formula, 10 samples were generated. Nine out of ten sample regression models agree with the population model, with positive coefficients for visibility and wind direction and negative values for humidity and wind speed. There is an abnormality in sample #4. While the other nine sample regression models are consistent with positive R²values, Sample #1 displays an oddity with negative values. The RMSE interval values for each regression model in this study fall between 4.334 and 9.582.

Performance Comparison of Support Vector Machine (SVM) and k-Nearest Neighbors (kNN) in Verifying Material Orientation

Eldio Utama, Eko Rudiawan Jamzuri — Sun, 22 Jun 2025 15:25:13 +0100

In automated manufacturing, verifying material orientation is essential to ensure the product assembly proceeds without errors. For instance, in the beverage industry, incorrect orientation of materials, such as bottle caps, can lead to failures in the packaging process, resulting in improperly sealed bottles that may compromise product quality and safety. This study compares the performance of Support Vector Machine (SVM) and k-Nearest Neighbors algorithms for verifying material orientation verification through automated optical inspection. The images were processed using the Inception V3 Convolutional Neural Network (CNN) to extract relevant image features, which were then classified using SVM and kNN algorithms. As a result, SVM achieved high classification performance during testing, with classification accuracy, precision, recall, and F1 score of 1.0 compared to kNN, which achieved only 0.967. However, kNN demonstrated superior computational efficiency, with a training time of 1.126 seconds and a validation time of 0.713 seconds, compared to SVM's training time of 3.101 seconds and validation time of 1.479 seconds. These results indicate that while both methods are highly effective for material orientation verification, kNN offers significant advantages in terms of computational speed, making it more suitable for real-time applications. The implications of this study highlight the potential for integrating the proposed method in industrial applications, promoting enhanced efficiency and reducing error rates in automated assembly lines.

Analisis Penerapan Mutual Information pada Klasifikasi Status Studi Mahasiswa Menggunakan Naïve Bayes

Sulfayanti Situju, Nahya Nur, Nursan Halal — Sun, 22 Jun 2025 15:32:23 +0100

Early identification of Student Study Status is essential for higher education institutions to implement proactive and strategic measures that facilitate timely completion of studies and mitigate dropout rates. This research intends to predict student study status with the Naïve Bayes method based on the features obtained from the implementation of Mutual Information. Feature selection through Mutual Information seeks to analyse the factors that most significantly impact the classification of student study status. The study status is categorized into three classes: dropout, enrolled, and graduate, based on 36 factors. The Mutual Information approach is employed to diminish data dimensions by discarding less relevant features while preserving critical information based on score values to achieve enhanced predictive accuracy. The selection of appropriate attributes enables the model to maintain simplicity while incorporating critical information aspects that significantly impact performance. Experiments were performed on a dataset comprising student academic variables, with data partitioning ratios of 80:20, 70:30, and 50:50 for training and testing datasets. The classification outcomes utilizing Naïve Bayes, without the use of Mutual Information across the three testing ratios, exhibited the accuracy of 68.29% in the 70:30 data split. Simultaneously, the classification outcomes utilizing Mutual Information across three test ratios are as follows: 71.64% accuracy at an 80:20 ratio with 10 selected attributes, 72.06% at a 70:30 ratio with 10 selected attributes, and the highest accuracy of 72.65% at a 50:50 ratio using 15 attributes. The utilization of the Naïve Bayes method for classifying student study status demonstrates enhanced accuracy when integrated with Mutual Information for feature selection. The findings of this study demonstrate that Mutual Information can streamline data by considering the quantity of attribute selections according to the ranking of their score values.

Penentuan Faktor Pemicu Gejala Penyakit Mata Glaukoma, Astigmatis, Hipermetropi, dan Miopi

I Kadek Arta Wiguna, Dewa Gede Hendra Divayana, Gede Indrawan — Sun, 22 Jun 2025 16:00:14 +0100

Expert systems support medical problem-solving, including the analysis of eye diseases. According to BPS RI (2022), over 8 million Indonesians suffer from visual impairments. In diagnosis, doctors often struggle to identify the primary causes of symptoms, impacting treatment effectiveness. This study proposes a system that combines Backward Chaining and Simple Additive Weighting (SAW) to systematically identify and prioritize causal factors of eye diseases. Backward Chaining is used to trace relationships between symptoms and the causes of glaucoma, astigmatism, hyperopia, and myopia. SAW is applied to assign weights to each causal factor and determine priority based on score ranking. Testing with 45 patient cases shows the system achieves 91% accuracy in identifying dominant causes. The 9% error rate stems from data limitations, subjective weighting in SAW, and inference rules in Backward Chaining. This system offers valuable support in early decision-making by helping doctors prioritize handling strategies based on the most significant underlying factors, thereby enhancing diagnostic efficiency and consistency.

News Classification using Natural Language Processing with TF-IDF and Multinomial Naïve Bayes

Nadira Alifia Ionendri, Feri Candra, Afdi Rizal — Tue, 24 Jun 2025 09:24:35 +0100

Online news contains valuable insights into public phenomena that can support statistical analysis by institutions like BPS Riau. However, current methods of classifying news are manual, time-consuming, and prone to human error. This study proposes an automated news classification system using Natural Language Processing (NLP) techniques with Term Frequency–Inverse Document Frequency (TF-IDF) for feature extraction and the Multinomial Naïve Bayes algorithm for classification. The dataset was collected via web scraping and manually labeled across five statistical categories: poverty, unemployment, democracy, inflation, and economic growth. The system achieved a validation accuracy of 83%, a test accuracy of 90%, with an average precision of 0.85, recall of 0.93, and f1-score of 0.87. These results demonstrate that the proposed system can significantly reduce the manual workload of news classification and be practically implemented by BPS Riau to support accurate and timely statistical reporting.