IKADO E-Journal (Institut Informatika Indonesia)
Not a member yet
470 research outputs found
Sort by
Comparative Analysis of Machine Learning and Deep Learning Models for PM2.5 and PM10 Time Series Forecasting Using the SISANAPAS Web Platform
No full textParticulate matter (PM2.5 and PM10) pollution remains a significant environmental concern in Indonesia. This study employed the SISANAPAS web platform to compare machine learning (ML) and deep learning (DL) algorithms for PM2.5 and PM10 forecasting. Using 3-hourly data from Cibeureum (January-October 2024), which underwent comprehensive pre-processing (K-Nearest Neighbors imputation, Z-score outlier removal, Min-Max scaling, feature engineering), Long Short-Term Memory (LSTM), Gated Recurrent Unit (GRU), Random Forest, and XGBoost models were evaluated. XGBoost Regression provided the most accurate forecasts, with an R² of 0.85, RMSE of 5.05 µg/m³, and MAE of 3.19 µg/m³ for PM2.5, and an R² of 0.85, RMSE of 9.71 µg/m³, and MAE of 6.55 µg/m³ for PM10. These results, significantly outperforming LSTM and GRU, highlight XGBoost\u27s potential for reliable air quality prediction and demonstrate SISANAPAS as a valuable tool for environmental data analysis, crucial for informing public health and environmental policies
Implementation of Blockchain Technology in E-Voting Using Smart Contract and ZK-SNARK
No full textE-voting systems are prone to challenges such as lack of transparency, risks of data manipulation, and dependence on centralized authorities, which can undermine trust in electoral processes. This research develops a blockchain-based e-voting system on the Polygon network, leveraging smart contracts and Zero-Knowledge Succinct Non-Interactive Argument of Knowledge (ZK-SNARK) to enhance security, transparency, and voter anonymity. The study employs an application development approach, implementing a structured methodology with initialization, registration, voting, and tallying phases. Smart contracts automate voter verification, vote casting, and result tabulation, while ZK-SNARK ensures voters can cast ballots anonymously without revealing their identities. The system’s transparency and immutability are tested using PolygonScan, demonstrating effective prevention of manipulations like double voting through cryptographic credentials (nullifier, commitment, and nullifier hash) and Merkle Tree structures. Results indicate that the system provides a secure, verifiable, and decentralized framework for elections. This implementation offers a robust foundation for future e-voting systems, promoting trust and integrity in digital voting processes
Human Computer Interaction Design in Web Based Warehouse Management Systems Using the Task Centered System Design (TCSD) Method
No full textIn the digital transformation era, web-based warehouse management systems play an essential role in ensuring operational efficiency, yet their effectiveness depends not only on functional accuracy but also on a user-centered interface design. This study aims to redesign the interface of Company XYZ’s warehouse management system using the Task-Centered System Design (TCSD) method within the Human-Computer Interaction (HCI) framework. TCSD emphasizes mapping real user tasks, analyzing user-centered requirements, prototyping through scenarios, and conducting usability evaluations. Data were collected through direct observation, structured interviews, and literature review, followed by interface prototyping in Figma and testing with the Maze platform. Six main warehouse tasks login, item distribution, returns, receiving, cycle counting, and searching were evaluated with five participants. The results demonstrated a Task Success Rate (TSR) of 100% across all tasks, with the shortest average Time on Task (ToT) for login (7.89 seconds) and item search (8.58 seconds), and the longest for cycle counting (58.10 seconds). Click Rate (CR) and Error Rate (ER) varied depending on task complexity, with login and item search showing higher ER values, which were influenced by predefined task flows. Overall, findings confirm that TCSD yields a practical and measurable design aligned with user workflows, producing an interface that is efficient, user-friendly, and adaptable to warehouse operations. This research contributes to HCI studies by demonstrating the effectiveness of scenario-based design in logistics systems, while also highlighting the need for larger-scale usability testing in future work
Pengembangan Daur Ulang Felt Modular Dari Limbah Kain Poliester Melalui Eksplorasi Motif dan Karakteristik Material
No full textPengelolaan sampah kain oleh Pemerintah belum optimal. Menurut data Sistem Informasi Pengelolaan sampah Nasional tahun 2025, menginformasikan bahwa pengelolaan sampah tekstil oleh Pemerintah sejumlah 32,36% yang dapat terkelola. Sedangkan 67,64% sampah belum terkelola dengan baik. Ada kemungkinan sampah kain sejumlah 2,64% menjadi salah satu komponen sampah yang belum terkelola secara optimal. Oleh karena itu penelitian pengembangan daur ulang felt modular dari limbah kain poliester sangat tepat untuk mengatasi permasalahan sampah kain yang dirasa belum terkelola dengan baik. Selain itu sampah kain terkait dengan isu lingkungan karena sebagian besar kain di pasaran mengandung poliester yang sulit terurai di alam sehingga mengganggu ekosistem lingkungan. Tujuan penelitian ini adalah mengembangkan teknik kempa dan eksplorasi motif menjadi felt modular yang estetik dan memiliki karakteristik material. Teknik pembuatan kain felt modular menggunakan medium lem tekstil sebagai perekat. Pendekatan penelitan Pengembangan daur ulang kain poliester ini Research and Depelovment meliputi eksplorasi dan eksperimentasi pengembangan teknik felt, eksperimen pada pengembangan. Hasil penelitian ini diharapkan dapat mengurangi dampak lingkungan, memberikan kontribusi inovasi material daur ulang terutama pengembangan teknik felt dan motif dari limbah kain poliester, serta menjadi panduan untuk desain modular pakaian
Development of Interactive Learning Application for Basic Programming Based on Technological Pedagogical Content Knowledge Framework
Full text linkInformation technology students must take Algorithms and Programming. Research shows that 28% of US students fail their basic programming subject, which is essential to mastering programming. In line with the previous study, 39% of students in the Informatics Engineering department’s basic programming course at campus X in the odd semester of 2022/2023 failed the course. The learning process should be able to integrate technology into it. An interactive learning application was developed utilizing the Technological Pedagogical and Content Knowledge (TPACK) framework, incorporating a pedagogical paradigm in its design through simulation elements and animated visuals. Through an extensive design, this learning application enhances student engagement by 78.3%, encouraging continued utilization in their educational process. The trial involving the group of students utilizing this application revealed that 5 out of 34 students failed the course, in contrast to 7 out of 33 students from the group that studied without the application
Single Sign-On (SSO) Implementation Using Keycloak, RADIUS, LDAP, and PacketFence for Network Access
Full text linkThe increasing demand for secure, seamless authentication mechanisms in public and private networks has fueled the need for more robust network access control (NAC) systems, as well as Single Sign-On (SSO) which is critical for organizations that require seamless and secure access across different platforms. This paper explores SSO in a fully open source implementations with Keycloak, RADIUS and LDAP; extending to captive portal implementations with PacketFence for Wi-Fi authentication. Specifically, this paper highlights the integration of PacketFence with FreeRADIUS for captive portal authentication, leveraging Keycloak for identity management and providing users with secure Wi-Fi access. Real-world examples, such as authenticating campus network users over Wi-Fi with 802.1X and captive portals, illustrate how these systems work in tandem to provide scalable and secure network access control. Testing showed up to 500 concurrent users with stable performance, minimal latency at a case study university. Key performance metrics included response times below 30ms
Factors Influencing Continuance Intention to Play Online Games
Full text linkThis study aims to identify and analyze the factors influencing continuance intention in playing online games, a rapidly growing global entertainment sector that attracts players from diverse backgrounds. Using the Expectation Confirmation Model (ECM), this research explores the psychological and behavioral aspects that drive players to continue playing. To achieve this objective, data were collected through a survey using Google Forms distributed to schools, colleges, and social media. A total of 505 active player responses were collected, and 469 valid data entries were retained after screening. The analysis was conducted using Structural Equation Modeling (SEM) with SPSS and AMOS software to identify the impact of each factor. The results from SPSS and AMOS calculations showed that Flow was not significant, and Engagement was excluded due to failing the validity test. These findings help developers and policymakers better understand player motivations to create more effective strategies for building a sustainable gaming industry. The study found that Social Influence had the greatest impact on continuance intention. Players were more likely to continue playing if the game was popular in their environment or had a large market. This factor fosters a sense of community and social support among players, from friends, family, and communities who play the same game. Perceived Usefulness and Perceived Enjoyment followed, contributing significantly as well
Ancient Javanese Manuscript Reconstruction Using Generative Adversarial Network with StarGAN v2 Variations
Full text linkAncient Javanese manuscripts are part of Indonesia\u27s cultural heritage; most of them are usually in bad condition due to the age and environmental surroundings. This paper presents a manuscript reconstruction using the Generative Adversarial Network model, using the variation of StarGAN v2. The primary objective of this research is to assist philologists in reconstructing damaged manuscripts more efficiently, reducing the time and effort compared to manual reconstruction methods. The training for 100 epochs is performed by the model in order to generate the reconstruction image closest to ground truth. This study is done on a dataset that consists of a set of damaged manuscript images. In this dataset, 80% is for training, 20% is for validation, and 10 images are used for testing. Quality assessment will be made on image outputs during training, based on PSNR, SSIM, and LPIPS metrics. The results indicate that the PSNR increases from 16.1234 dB at the 50th epoch to 17.5588 dB at the 100th epoch, while the SSIM increases from 0.8374 to 0.8519, showing a strong improvement in image quality. Despite the LPIPS having a very slight increase from 0.1020 to 0.1051, this evidences that the model can be further improved. Overall, this study demonstrates that the StarGAN v2 model is effective in reconstructing ancient Javanese manuscripts-a great contribution to the field of cultural heritage preservation using modern technology
Comparative Analysis of Naïve Bayes Algorithm Performance in English and Indonesian Text Sentiment Classification on Duolingo Application in Playstore
Full text linkText classification is an important topic in Natural Language Processing (NLP), especially when conducting research on user reviews on language learning apps such as Duolingo. This study compares the effectiveness of the Naïve Bayes algorithm in identifying sentiment in English and Indonesian reviews on the Duolingo app on Playstore. The approach includes data collection, text preparation (case folding, tokenization, stopword removal, and stemming), and Naïve Bayes algorithm evaluation for each dataset. Model performance was evaluated using accuracy, precision, recall, and F1-score. The Naïve Bayes method obtained 84% accuracy on the English dataset with a 90:10 data split and 67% accuracy on the Indonesian dataset with the same split ratio. The difference in the results obtained is due to several variables, including the use of informal language, slang, and more complicated word variants in Indonesian, which make proper classification more difficult for the model to achieve
Classification of Anxiety Levels of IGD Patients at RSU Royal Prima Medan Using Support Vector Machine (SVM) Algorithm
No full textThe level of patient anxiety in the Emergency Department (ED) is an important indicator that affects the diagnosis and medical management process. However, the classification of anxiety levels is often hampered by data imbalance, which can reduce the accuracy of predictive models. This study aims to develop a patient anxiety level classification model in the ED using the Support Vector Machine (SVM) algorithm with the application of the Synthetic Minority Oversampling Technique (SMOTE) to address the class imbalance issue. The data used consists of 734 patient samples divided into 80% training data and 20% testing data, including physiological parameters such as systolic and diastolic blood pressure, respiratory rate, heart rate, and demographic data such as age and gender. The preprocessing process includes imputing missing values and normalizing numerical features so that the model can learn optimally. Performance evaluation shows that the use of SMOTE increases classification accuracy from 95% to 97%, as well as improving precision, recall, and F1-score metrics at almost all anxiety levels. Visualization of the relationships between numerical features also reveals significant correlation patterns between physiological variables and patient anxiety levels. The results of this study confirm the effectiveness of SMOTE in addressing data imbalance, thus producing a more accurate anxiety level classification model that can serve as an aid in clinical decision-making. Thus, the developed model has significant clinical utility potential as a diagnostic aid that can accelerate and improve the accuracy of patient management in the emergency department, thereby supporting the overall improvement of healthcare service quality