Journal of System and Computer Engineering

Application of Advanced Encryption Standard (AES) Algorithm in E-Commerce Login System for User Data Security

2025-01-25T01:32:07+00:00

E-commerce becomes an electronic media that uses a login system used by users. User data in the form of usernames and passwords is vulnerable to hacking. One technique to improve user security is the implementation of AES algorithms on login systems in E-Commerce applications. The purpose of this study is to apply the AES algorithm in the login system of e-commerce websites and analyze the improvement of information security for users after the implementation is carried out. The research method used is an experiment with the application of the use of the AES algorithm before and after. Therefore, the application of the AES algorithm on the login system of e-commerce websites can be used as a solution to improve user data security. Testing using Wireshark and Burpsuite tools. The results obtained are that AES successfully secures the username and password on the e-commerce login system .

Home-based Waste Monitoring System using Internet of Things with Fuzzy Logic Method

2025-01-25T01:34:16+00:00

The accumulation of waste in certain locations, especially in residential areas, due to the continuous accumulation of waste can cause environmental disturbances such as disease and unpleasant odors. This system is designed to find out whether the trash bin is full or not by applying fuzzy logic, if the status of the trash bin can be cleaned, it will be handled immediately so that the waste does not accumulate and disturb people around. This system can find out when the last time the waste was taken and the location of the trash bin as a prototype which can later be applied to areas with a wider range by cleaning service officers in Mamuju City. This system uses the HC-SR04 sensor to detect the distance of the waste to the sensor and uses the Load Cell sensor to detect the weight of the waste. Several tests were carried out, first by measuring the accuracy of each sensor used, the HC-SR04 sensor accuracy was obtained at 96.68% with an error of 3.32%. While the accuracy of the load cell sensor is 90.68% with an error of 9.32%. The second test calculates the sensor response time and Blynk notification response since the sensor detects waste, the average HC-SR04 sensor detection response is around 0.83 seconds. For the response time of incoming notifications when there is movement in the HC-SR04 sensor area has an average of 2.65 seconds. While the Load Cell sensor response time is only around 0.53 seconds. For Blynk notification response time since the Load Cell sensor detects it has an average of 2.51 seconds. The third test calculates the response time of the two sensors (HC-SR04 and Load Cell) and the Blynk notification response since the two sensors detected the waste, the average response time of the two sensors finished detecting only around 0.91 seconds. For the response time of the trash bin status condition, if there is movement in the area of the two sensors, the condition of the trash bin will change and display the status of Normal, Needs Cleaning and Highly Needs Cleaning with an average response time of 1.18 seconds. The system successfully sends notifications according to the fuzzy rules and expected to speed up the waste handling process

Clusterization Of Infant Data Based on Posyandu Examination Using K-Means

2025-01-25T01:37:36+00:00

Nutrition is a very important aspect for the human body, especially in toddlers and children. Balanced nutrition not only supports children's growth and development, but also improves academic achievement and contributes positively to their future development. However, in Tumpiling Village, the problem faced is the low basic understanding of parents and Posyandu cadres regarding balanced nutrition in early childhood. This causes toddlers to still be found with malnutrition or obesity, as well as a lack of data collected based on children's nutritional characteristics. Clustering, as one of the popular methods in processing medical, biometric, and various other fields of data, is known for its simplicity and effectiveness in grouping large-scale data based on similar characteristic. This study aims to groups the nutritional status of toddlers based on height and weight parameters using the K-Means Clusterings algorithm. This grouping produces several categories of nutritional status, namely obesity, overnutrition, good nutrition, undernutrition, and poor nutrition. By applying the Clustering method using K-Means, the nutritional status of toddlers can be classified more clearly, so that it can be a basis for Posyandu cadres in taking early preventive measures against malnutrition and obesity. In this study, the author used 28 toddler data. From the data, the author randomly determined the cluster center of 5 data, which then resulted in the following grouping: 7 toddlers experienced malnutrition, 3 toddlers were undernourished, 6 toddlers with good nutrition, 7 toddlers were overnourished, and 5 toddlers were obese. These results indicate the need for further attention and action from Posyandu and Puskesmas cadres to help parents in overcoming toddler nutrition problems

Facial Expression Recognition of Al-Qur'an Memorization Students Using Convolutional Neural Network

2025-01-25T01:40:23+00:00

Facial expression recognition technology has advanced significantly and has become an intriguing topic of study. This research focuses on the facial expressions of Al-Qur’an memorization students, which naturally reveal various aspects of their engagement, understanding, and emotional barriers about the verses being memorized. The issue is that facial expression recognition still lacks optimal accuracy, and the need for a better algorithmic model to improve accuracy is evident. Therefore, an intelligent computing system is required to address this problem. This study aims to enhance the accuracy of facial expression recognition in Al-Qur’an memorization students using the Convolutional Neural Network (CNN) method, classifying facial expressions such as happy, neutral, and tired based on collected facial image data, achieving improved accuracy. The first stage involves capturing image data via CCTV, followed by preprocessing, training the CNN model, result analysis, and model evaluation. By using the CNN method to recognize the facial expressions of Al-Qur’an memorization students, a high accuracy of 84% was achieved with a loss value of 14.9.

Automated Medical Image Processing for Lung Pneumonia Diagnosis Based on LS-SVM

2025-01-25T01:42:24+00:00

Pneumonia is an inflammation of the lungs that causes pain when breathing and limits oxygen intake. Pneumonia can be caused by bacteria, viruses, and fungi. Image processing, a branch of informatics or computer science, is a field highly related to the manipulation and analysis of digital images. This study aims to design a medical image processing system as an alternative to support the diagnosis of Pneumonia in the lungs using the LS-SVM method. LS-SVM (Least Square Support Vector Machine) is a simpler and modified model of the SVM method. HoG (Histogram of Gradient) is a commonly used feature extraction method in image processing and object detection. The objective of this study is to improve the quality of healthcare services and assist in faster and more accurate clinical decision-making. The results show that lung image analysis using the LS-SVM method has a good accuracy level in the image classification process, with 2000 training data inputs processed in the preprocessing stage, consisting of 1000 Pneumonia images and 1000 normal lung images, while the testing data used consisted of 500 images, with 250 Pneumonia images and 250 normal lung images. Based on the tested data, the system achieved an accuracy of 81% for 1300 tests, proving that the LS-SVM method is effective in image processing with satisfactory results.

Air Conditioner Control and Monitoring System based on Temperature Balance in Server Room using Fuzzy Logic and Internet of Things Methods

2025-01-25T01:45:05+00:00

This research develops a temperature and humidity control system in the server room based on the Internet of Things and using fuzzy logic algorithms at AMIK Luwuk Banggai. The system is designed using NodeMCU ESP32, DHT11 sensor, Arduino IDE, and Blynk application, with objective of monitoring and controlling environmental conditions in real time. A series of quantitative experiments were conducted to evaluate the effectiveness of the sensor system. These experiments involved observations, measurements, and a comparison of the results with manual calculations. The results demonstrate that the DHT11 sensor exhibits a margin of error of 1.21% and a hardware accuracy rate of 98.79%. Furthermore, the integration of the Internet of Things (IoT) and the implementation of fuzzy logic in air conditioner control studies, as demonstrated in this study, has the potential to enhance the accuracy of temperature and humidity control within the room server to an accuracy rate of 90.91%. Furthermore, it can improve the responsiveness of the system in maintaining temperature stability. These findings were observed at AMIK Luwuk Banggai, where the application of IoT and fuzzy logic has been implemented. Fuzzy logic offers an effective and dependable approach to regulating temperature fluctuations in the server room, ensuring a stable environment that minimizes the likelihood of operational issues or hardware damage. The objective is to extend the lifespan of the hardware by preventing such complications.

Prototype of Internet of Things (IoT) and Android-based Chilli Plant Watering Monitoring and Control Tool

2025-01-25T01:47:08+00:00

Chili is a very important vegetable commodity in everyday life. The purpose of the researcher is to design a prototype of a chili plant watering monitoring and control tool based on the Internet of Things (IoT) and Blynk to make it easier for chili farmers to monitor and control the watering of chili plants remotely. The research method uses Research and Development (R&D), which aims to produce certain products and test their effectiveness. The results of this tool prototype use acrylic as a container to unite various tool components and other supporting components. The soil moisture sensor is used to detect soil moisture. The DHT11 sensor is used to detect air humidity. The DS 18B20 sensor is used to detect soil temperature. The PIR Motion sensor is used to detect objects. Information about the measurement results on the sensor will be displayed on the LCD screen and the Blynk application. DC water pump is used to remove water from the container and spray it on the plants. The test results of the Prototype concluded that the process of watering is automatic when the measurement results of the soil moisture Threshold are SP Low 40%, SP High 60%, and Temperature High 31^o C, meaning that the watering process occurs when the humidity is <40% and the soil temperature <31^oC and stops when the humidity is > 60%.

Automatic Bird Pest Repellent System in Rice Farming Land

2025-01-25T01:49:43+00:00

The importance of advancing sustainable rice farming is because rice is a staple food whose production must be increased to meet the basic needs of the community. The problem that occurs to rice farmers is that thousands of bird pests look for food in fertile rice fields, causing great losses for farmers. The way farmers efficiently repel bird pests is by developing technology that can monitor bird activity around rice fields in real time remotely and is able to repel bird pests by using the Internet Of Things (IoT) development. which aims to design IoT or tools that can monitor bird activity in real time around agricultural areas remotely, by utilizing PIR sensors to detect bird pests, LDR resistors detect light, ESP32-CAM to monitor rice fields and utilizing solar panel power as a sound signal generator that will sound disturbing the bird's hearing system so that the birds fly away and move the DC motor to repel bird pests automatically and the data collected from the ESP32-CAM (1) sensor will be sent to the IoT platform with the Telegram application connected via an internet connection, allowing farmers to monitor remotely via smart devices such as smartphones or computers.

Electronic Equipment Power Usage Control and Monitoring System in the Home Internet Of Things (IOT) Based

2025-01-25T01:51:58+00:00

The objectives of this study are (1) to achieve energy efficiency and cost savings (2) Internet of Things (IoT)-based control and monitoring systems using Node MCU ESP32 as a data processing center (3) enabling data processing from PZEM-004T sensors and sending control commands to solid state relays (SSR) based on user input via a website application. The implementation of this system shows significant potential in reducing energy consumption and costs in households. With real-time feedback on energy consumption, users can make wiser decisions about the use of electronic equipment, thereby reducing energy waste. Remote control capabilities allow users to manage electronic equipment more effectively, improve security, and reduce unnecessary energy consumption. This study shows that manual electricity usage reaches 9.59%, while with the implementation of the IoT system it is only 5.49%, so there is a saving in electricity consumption of 4.1%. This proves that the IoT system is more effective and efficient in managing the power consumption of electronic equipment.

Hill Cipher-Based Visual Cryptography for Copyright Protection of Images Using Flexible Matrix Keys

2025-01-25T01:52:55+00:00

The widespread distribution of digital images on the internet has diminished the copyright protection associated with them. In some cases, copyrighted and economically valuable digital images should not be modified or distributed without permission, as altering the original image can harm its owner. This violation is common, but many internet users are unaware of it. The goal of this research is to protect intellectual property rights of digital images using visual cryptography based on the Hill Cipher algorithm with matrix key flexibility. Hill Cipher is chosen for its ability to encrypt data in blocks, making it more secure than classical cryptographic algorithms that encrypt data individually. Visual cryptography is used to secure digital images through encryption and decryption. Encryption scrambles the image, while decryption restores it. The research method involves collecting digital image datasets, preprocessing, Hill Cipher encryption, and decryption. Key flexibility includes matrix keys of 2x2, 3x3, and 4x4 to enhance security. This research has demonstrated the effectiveness of the Hill Cipher algorithm in protecting digital images through encryption and decryption processes with flexible matrix keys of size 2x2 and 3x3. The results of the experiments, including encryption and decryption using both matrix sizes, have been thoroughly analyzed with respect to various cryptographic metrics: histogram analysis, energy, entropy, and running time.

Recognition of Human Activities via SSAE Algorithm: Implementing Stacked Sparse Autoencoder

2025-01-25T01:54:13+00:00

This study evaluates the performance of Stacked Sparse Autoencoder (SSAE) combined with Support Vector Machine (SVM) against a standard SVM for classification tasks. We assessed both models using accuracy, precision, sensitivity, and F1 score. The SSAE Support Vector Machine significantly outperformed the standard SVM, achieving an accuracy of 89% compared to 37%. SSAE also achieved higher precision (87% vs. 75%) and sensitivity (89% vs. 37%), with an F1 score of 88% versus 36% for the standard SVM. These results indicate that SSAE enhances the model’s ability to capture complex patterns and provide reliable predictions. This study highlights the effectiveness of SSAE in improving classification performance, suggesting further research with larger datasets and additional optimization techniques to maximize model efficiency

Implementation of an Internet of Things (IoT)-Based Air Quality Monitoring System for Enhancing Indoor Environments

2025-01-25T01:54:55+00:00

This research investigates the development and implementation of an IoT-based air quality monitoring system designed to improve indoor environmental conditions. The primary objective of this study is to develop a comprehensive system that continuously monitors air quality parameters, including smoke, LPG gas, carbon monoxide (CO), temperature, and humidity. The system integrates real-time data collection from various sensors, which is then processed and transmitted to a cloud platform for secure storage and detailed analysis. The user-friendly interface of the software allows for intuitive monitoring and reporting, while built-in notification and alert features ensure timely responses to significant air quality changes. Testing results demonstrate that the system operates with high reliability, providing accurate data and stable performance. The findings confirm that the system effectively addresses indoor air quality concerns and offers valuable insights for maintaining a healthy and safe environment. This research contributes to the field by showcasing a practical application of IoT technology in environmental monitoring.

ARIMA Method Implementation for Electricity Demand Forecasting with MAPE Evaluation

2025-01-25T01:55:55+00:00

Electricity demand forecasting is critical for efficient energy management and planning. This study focuses on the development and implementation of the Autoregressive Integrated Moving Average (ARIMA) method for forecasting electricity demand in South Sulawesi's power system. The evaluation of forecasting accuracy was conducted using the Mean Absolute Percentage Error (MAPE), which measures the percentage error between predicted and actual values. Two experiments were conducted with different ARIMA models: ARIMA(5,1,0) and ARIMA(2,0,1). Results showed that the ARIMA(5,1,0) model achieved a MAPE of 2.15%, while the ARIMA(2,0,1) model performed slightly better with a MAPE of 1.91%, indicating highly accurate predictions. The findings highlight the effectiveness of the ARIMA method in forecasting electricity demand, providing a reliable tool for energy providers to optimize resource allocation and enhance operational efficiency. Future research may explore integrating ARIMA with other advanced methods to further improve forecasting performance.

CNN Modeling for Classification of Bugis Traditional Cakes

2025-01-25T01:56:58+00:00

Abstract

This research aims to create a classification system that can recognize traditional Bugis cakes using the Convolutional Neural Network method. (CNN). Traditional Bugis cakes play an important role in Indonesia's culinary heritage, which is rich in diversity and flavor. However, the lack of documentation and sufficient recognition of these cakes could lead to the loss of cultural knowledge. In this study, a collection of images of traditional Bugis cakes was gathered and processed for training a CNN model. This model was created to recognize and classify various types of cakes based on their visual attributes. The evaluation results show that the CNN model can achieve a high level of accuracy in identifying these cakes, making it a useful tool in preserving and promoting traditional Bugis cakes. This research is expected to contribute to the development of image recognition technology and raise public awareness about the richness of local culinary heritage.

Keywords : Convolutional Neural Network (CNN), Bugis Cake, Indonesian Cuisine

Parameter Optimization Supports Vector Machine Using Genetic Algorithms to Improve the Efficiency of Data Transfer Prediction on Google Cloud

2025-01-25T01:57:51+00:00

Efisiensi transfer data merupakan elemen kunci dalam infrastruktur cloud seperti Google Cloud. Penelitian ini bertujuan untuk meningkatkan akurasi prediksi efisiensi transfer data menggunakan Support Vector Machine (SVM) yang dioptimasi dengan Algoritma Genetika (GA). Dataset berisi informasi tentang ukuran file, latensi jaringan, utilisasi server, dan waktu transfer data. Algoritma Genetika diterapkan untuk mencari parameter optimal, yaitu nilai C dan gamma. Hasil penelitian menunjukkan bahwa optimasi parameter menggunakan GA mampu meningkatkan akurasi prediksi hingga 90%, dibandingkan metode tradisional Grid Search yang mencapai akurasi maksimal sebesar 88%.

Evaluating the Effectiveness of Online Learning Methods with a Probabilistic Naive Bayes Approach

2025-01-25T01:58:43+00:00

Online learning methods become an important element in supporting the flexibility and effectiveness of teaching and learning process, especially through approaches such as Video Tutorial, Virtual Discussion, and Self-paced Reading. This research aims to evaluate the effectiveness of the three methods in improving students' engagement, comprehension, and learning motivation by utilizing Naive Bayes algorithm. The dataset used includes student data taken through questionnaires and teacher evaluation results, with variables such as material suitability, engagement, ease of access, and student exam results.

Through this approach, the research is able to predict the learning method that best suits students' needs based on the analyzed variables. The results show that Video Tutorial is the most effective method in supporting students' understanding and motivation. The implementation of this research is expected to help the development of a better online learning system in improving students' learning experience, and provide practical recommendations for educators in choosing the right learning method.

Performance Analysis of API in Google Cloud Storage Service Integration

2025-01-25T01:59:23+00:00

Google Cloud Storage (GCS) is one of the leading cloud storage services that supports large-scale data management through API integration. APIs allow applications to upload, download, and manage data in real-time. This study aims to analyze the performance of APIs in integration with GCS using response time, throughput, and latency parameters. Tests were conducted on various scenarios, including massive data transfer, distributed data management, and caching usage. The results showed that the average API response time reached 120 ms under normal conditions and increased to 180 ms under high load. Throughput reached an average of 400 MB/s, but decreased when the number of simultaneous requests increased. The average server latency was recorded at 60 ms and can be optimized with caching technology. Implementation of strategies such as Content Delivery Network (CDN) and request header optimization can improve performance by up to 30%. This study provides practical guidance for developers to optimally utilize GCS APIs in large-scale data management.