Journal of Innovation Information Technology and Application (JINITA)

Performance Optimization in Three-Modality Biometric Verification using Heterogeneous CPU-GPU Computation

2025-01-04T16:28:10+07:00

This paper proposes a method to improve the performance of tri-modal biometric verification using a heterogeneous computing system exploiting the synergy between CPU and GPU. The main objective is to reduce the time required for verification while maintaining the system's accuracy. The design of this system is based on a decision fusion algorithm based on the logical OR connector, enabling the results of the three modalities to be combined. The implementation is being carried out in C# with Visual Studio 2019, using the Task Parallel Library to parallelize tasks on the CPU, and OpenCL.NET to manage processing on the GPU. The tests carried out on a representative sample of 1,000 individuals, show a clear improvement in performance compared with a sequential system. Execution times were significantly reduced, ranging from 0.03 ms to 0.67 ms for data sizes between 50 and 1000. Analysis of the performance gains, based on Amdahl's law, reveals that the proportion of tasks that can be parallelized remains higher in heterogeneous systems than in parallel and sequential systems, even though part of processing remains sequential for large data sizes. This study highlights the ability of heterogeneous computing systems to effectively reduce the verification time of biometric systems while maintaining an optimal balance between processing speed and overall efficiency. The results demonstrate the potential of this approach for advanced biometric applications, particularly in distributed environments.

Comparative Analysis of Keypoint Detection Performance in SIFT Implementations on Small-Scale Image Datasets

2025-01-04T16:25:50+07:00

Scale-invariant feature transform (SIFT) is widely used as an image local feature extraction method because of its invariance to rotation, scale, and illumination change. SIFT has been implemented in different program libraries. However, studies that analyze the performance of SIFT implementations have not been conducted. This study examines the keypoint extraction of three well-known SIFT libraries, i.e., David Lowe's implementation, OpenSIFT, and vlSIFT in vlfeat. Performance analysis was conducted on multiclass small-scale image datasets to capture the sensitivity of keypoint detection. Although libraries are based on the same algorithm, their performance differs slightly. Regarding execution time and the average number of keypoints detected in each image, vlSIFT outperforms David Lowe’s library and OpenSIFT.

Sentiment Analysis of Universitas Jember’s Sister for Student Application Using Gaussian Naive Bayes and N-Gram

2025-01-04T16:24:03+07:00

This research aims to classify sentiment in reviews of the Universitas Jember Sister for Student application on Google Play Store, a vital student platform. The primary challenge tackled is the automated identification of positive and negative user sentiments. The study employs the Gaussian Naive Bayes method for classification and uses N-Gram techniques for sentiment analysis. The dataset consists of 1097 reviews, with 673 negative and 424 positive reviews, after removing 86 neutral spam reviews. The data is divided into 80% training data (877 reviews) and 20% test data (220 reviews). Gaussian Naive Bayes is used for modeling and combined with TF-IDF vectorization. The findings reveal that the Gaussian Naive Bayes model achieves an accuracy of 68%, precision of 68%, and recall of 71% on the test data. N-Gram analysis shows frequent occurrences of words like "bisa", "bagus", and "aplikasi" in positive sentiments, while "bisa", "hp", and "absen" are prevalent in negative sentiments. The study concludes that the Gaussian Naive Bayes model effectively classifies sentiment in application reviews, with the potential for further performance improvements.

Mitigating the Risks of Enterprise Software Upgrades: A Change Management Perspective

2025-01-04T16:22:38+07:00

Enterprise software upgrades are crucial for maintaining competitive advantage, ensuring security, and enhancing operational efficiency. However, these upgrades often pose significant risks, including system disruptions, data loss, and user resistance. The problem lies in effectively managing these risks to avoid operational setbacks and ensure successful adoption. This paper explores the role of change management in mitigating these risks by offering solutions through strategic planning, stakeholder engagement, and comprehensive training programs. The research employs a mixed-methods approach, integrating quantitative survey results from 185 participants and qualitative insights from 20 in-depth interviews. Results indicate that organizations prioritizing stakeholder engagement, tailored training, and proactive communication achieve higher user satisfaction, improved system performance, and enhanced operational efficiency. These findings provide a framework for best practices in change management that minimize risks and promote successful software upgrades.

Application of Machine Learning in Fault Detection And Classification in Power Transmission Lines

2025-01-04T16:17:35+07:00

Electrical faults have been identified as a significant contributing factor to electrical equipment damage. Such incidents can potentially result in a range of adverse consequences, including bushfires, electrical outages, and power shortages. The detection and classification of faults facilitates the delivery of superior quality of service, the preservation of the environment, the prevention of equipment damage, and the satisfaction of electricity line subscribers. In this study, we analyze the data from an electrical network comprising four generators of 11 kV, which have been modeled in Matlab. The generators are situated in pairs at either end of the transmission line. Subsequently, machine learning techniques are employed to detect faults in the transmission between lines, and machine learning models are utilized to classify the faults. Four distinct supervised machine learning classifiers are employed for comparison purposes, with the results presented in a confusion matrix. The results demonstrated that decision trees are particularly well-suited to this task, with an 88.6205% detection rate and a slightly higher accuracy than the random forest algorithm (87.9212% detection rate). The K-nearest neighbor's approach yielded a lower result (80.4196% of faults detected), while logistic regression demonstrated the lowest performance, with 34.5836% of faults detected. Six fault categories were found in the dataset: No-Fault (2365 occurrences), Line A Line B to Ground Fault (1134 occurrences), Three-Phase with Ground (1133 occurrences), Line-to-Line AB (1129 occurrences), Three-Phase (1096 occurrences) and finally Line-to-Line with Ground BC (1004 occurrences).

Development of Android-Based Interactive Learning Media on Computer Assembly Material with the ADDIE Model Approach

2025-01-04T16:17:01+07:00

This study raises the title of the development of android-based interactive learning media on computer assembly material at SMK Negeri 1 Lolayan. Based on the observations and interviews, several problems were found in learning computer assembly material, including teachers still using PowerPoint learning media that is only text-based. During the practicum, one of the computers used was damaged. This is due to limited knowledge and inadequate facilities, resulting in teachers' lack of innovation and creativity in developing learning media, which makes it difficult for students to understand the material. This study aims to develop android-based interactive learning media on Computer Assembly material for class X TKJ students at SMK Negeri 1 Lolayan and test the feasibility of interactive learning media through material and media expert feasibility tests and determine the practicality of learning media through respondent trials (students/users). The research method used is Research and Development (R&D) with the ADDIE development model (Analyze, Design, Development, Implementation, Evaluation). The results of this study were obtained from feasibility testing by material experts, who obtained an average value of ‘138’ with the criteria ‘Very Feasible.’ The results of feasibility testing by media experts obtained an average value of ‘120’ with the criteria ‘very Feasible,’ and the results of testing student response responses obtained an average value of ‘101’ with the criteria ‘Very Feasible’. The results showed that Android-based interactive learning media is feasible to use as an alternative to learning computer assembly

An Intelligent System for Light and Air Conditioner Control Using YOLOv8

2025-01-04T16:16:15+07:00

High energy consumption in classrooms is a significant concern, often resulting from inefficient lighting and air conditioning systems. Specifically, the problem lies in the lack of automated control mechanisms that adjust energy use based on real-time occupancy data. This study aims to develop and evaluate a system that employs a camera integrated with the YOLOv8 algorithm to detect human presence and optimize energy usage by controlling lights and air conditioning. The system's performance was assessed in three different classroom environments: two large and one small. The system's accuracy for occupancy detection varied from 13.64% to 100%, depending on lighting conditions and room size. Light control accuracy was highest in the classrooms with consistent lighting, reaching 99.77%. Air conditioning control achieved perfect accuracy of 100% in the classroom with a SHARP brand AC, with a maximum remote-control range of 7 meters. These findings indicate that the system's performance is influenced by lighting conditions and room size, with smaller rooms showing better results. The system demonstrates promising potential for reducing energy consumption in classroom settings, thereby contributing to more sustainable energy practices.

Programming Languages Prediction from Stack Overflow Questions Using Deep Learning

2025-01-04T16:15:44+07:00

Understanding programming languages is vital in the ever-evolving world of software development. With constant updates and the emergence of new languages, staying informed is essential for any programmer. Additionally, utilizing a tagging system for data storage is a widely accepted practice. In our study, queries were selected from a Stack Overflow dataset using random sampling. Then the tags were cleaned and separated the data into title, title + body, and body. After preprocessing, tokenizing, and padding the data, randomly split it into training and testing datasets. Then various deep learning models were applied such as Long Short-Term Memory, Bidirectional Long Short-Term Memory, Multilayer Perceptron, Convolutional Neural Network, Feedforward Neural Network, Gated Recurrent Unit, Recurrent Neural Network, Artificial Neural Network algorithms to the dataset in order to identify the programming languages from the tags. This study aims to assist in identifying the programming language from the question tags, which can help programmers better understand the problem or make it easier to understand other programming languages.

Mobile-Based Event Decoration Ordering System Using UAT Method with PIECES Framework

2025-01-04T16:14:54+07:00

The Mobile Event Decoration Booking System is an innovative solution designed to facilitate users in ordering event decorations. By implementing the User Acceptance Testing (UAT) method and the PIECES framework, this system ensures that the developed application meets the needs and expectations of users. This research aims to identify and analyze key features in the ordering process and evaluate user satisfaction with the application. Respondents provide valuable feedback regarding the interface, functionality, and overall user experience through UAT. The research results indicate that this application can enhance the efficiency of bookings, reduce communication errors between service providers and customers, and offer a better experience. With the application of the UAT method, users feel that this system effectively meets their needs, resulting in an improved experience in event planning. These findings suggest that the factors influencing user satisfaction and interest are adequate and should be maintained. The Mobile Event Decoration Booking System has successfully improved the efficiency and effectiveness of the booking service, with an average user satisfaction rate of 95%.

Website Security Analysis Using Vulnerability Assessment Method

2025-01-04T16:14:11+07:00

In today’s digital era, ensuring website security is crucial, especially in the education sector which is frequently targeted by cyber attacks. This research aims to test security of the Universitas Internasional Batam (UIB) website using OWASP ZAP and Nessus. The method will be used in this research was vulnerability assessment. It will involve gathering information with the tools such as, Nmap, whois and nslookup. OWASP ZAP detected 11 vulnerabilities, categorized into 6 medium level and 5 low level, including Content Security Policies (CSP) and anti-clickjacking headers. Otherwise, Nessus only detected one medium level vulnerability, the absence of HTTP Strict Transport Security (HSTS). The difference in detection results from the tools that OWASP ZAP is better at finding web application weakness that are consistent with the OWASP Top Ten 2021, while Nessus specifically targets server and network configuration. For educational institutions, these results emphasize the importance of conducting regular vulnerability assessment to protect sensitive data. Recommended action include implementing CSP to prevent Cross-site scripting (XSS) and other injection attacks, enforcing HSTS to secure communication, and its recommend to updating software to mitigate the unknown vulnerabilities. By adopting these measures, institutions can reduce their exposure to cyber attacks, its also can maintain user trust, and strengthen overall security. This research provides a pratical framework for stregthening the security of educational websites against evolving threats. These findings highlight that the importance of using multiple tools can provide a more comprehensive view of security gaps.

The Influence of the Tiktok Application on Cyberbullying Behavior

2025-01-04T16:12:37+07:00

Cyberbullying is threatening, insulting, or intimidating behavior carried out through online media. This cyberbullying behavior is vulnerable to being carried out or felt by teenagers who are still easily instigated by bad actions around them. Therefore, this study aims to determine what effects the TikTok application has on cyberbullying behavior in adolescents and to find out the causes and handling solutions for cyberbullying behavior. The research was conducted using the Technology Acceptance Model (TAM) method and the descriptive quantitative method. The research was conducted from June 11 to June 21, 2024, with a sample size of 91 students determined using the proportionate stratified random sampling method. The results of hypothesis testing with the t-test state that perceived usefulness has no effect on real conditions of use, then perceived ease of use and behavior to continue using positively affect real conditions of use. Meanwhile, attitude towards use harms the real conditions of use. The f-test states that all variables have a simultaneous effect. Meanwhile, the R-Square test states that perceived usefulness, perceived ease of use, attitude towards use, and behavior to continue using contribute 62.4% to the real conditions of use