JITSI : Jurnal Ilmiah Teknologi Sistem Informasi

Evaluation of Train Position Detection Using Radio Ranging in CBTC

2026-01-02T08:57:01+00:00

This study analyzes the performance of Radio Ranging technology in improving train position detection accuracy within the Communications-Based Train Control (CBTC) system of MRT Jakarta. The simulation-based experimental method was carried out using MATLAB, with the comparison of Radio Ranging and Tachogenerator (TG) across three operational conditions: open area, tunnel, and dense urban environment. Evaluation metrics include position deviation and Root Mean Square Error (RMSE). The results show that Radio Ranging provides better accuracy compared to TG, particularly in tunnel and dense areas, where TG tends to accumulate errors. These findings indicate that Radio Ranging can serve as a complementary method to enhance CBTC reliability and operational safety of MRT Jakarta.

Leveraging Blockchain and Federated Reinforcement Learning for Enhanced Fraud Detection in Financial Transactions

2026-01-02T08:57:20+00:00

Credit card fraud detection uses several technologies including machine learning (ML) and statistical analysis as well as card authentication methods. Credit card fraud is the illegal use of someone's credit card for purchase of goods or services. Comprehensive studies have shown how well ML technology generates exact prediction models to spot possible transaction fraud. Though the potential of losing the actual card is still a worry, hackers have been increasingly acquiring credit card numbers and personal information online. The growth in e-commerce has matched credit card use for online transactions, which has surged credit card theft. Complex detection systems including federated reinforcement learning (FRL) and blockchain technology have been developed to solve this challenge. Using standard pattern matching methods can make it difficult to tell real from fake transactions. A decentralised method of ML, FRL stresses user privacy and enhances anonymity and confidence in financial transactions. An original approach for teaching a credit card fraud detection model is shown in this work It makes advantage of user behaviour traits, federated reinforcement learning, and blockchain technologies. Using a smart contract between the bank and the client, the approach aims to reduce dishonest activity. By lowering the dependence on centralised data aggregation, this innovative approach guarantees user privacy protection all through model development. The research also looks at the challenges in credit card fraud detection and offers ideas for next developments.

Implementasi Logika Sistem Smart Door Lock Berbasis Arduino dengan Kontrol Aplikasi Mobile untuk Peningkatan Keamanan menggunakan metode Fuzzy Sugeno

2026-01-02T08:56:41+00:00

The development of technology in Internet of Things (IoT) devices has driven global innovation in home security systems, particularly in door lock mechanisms, as conventional locks are prone to being lost or duplicated. Most previous studies we reviewed designed home security lock systems using Raspberry Pi, ESP32, fingerprint sensors, RFID, and third-party platforms. Existing smart locks often fall short when it comes to affordability, user interface flexibility, and the quality of their mobile apps. Our project addresses these shortcomings by developing a new smart door lock prototype. The system consists of an ESP32-based unit controlled via a purpose-built mobile application. The real brains of the operation is the Fuzzy Sugeno method, which handles all the decision-making. The ESP32 is at the core of our system, responsible for interpreting instructions from the mobile application. We evaluated its effectiveness by testing several critical parameters, including access time, the system's response to failed attempts, and the validity of user credentials. The test results were a success: the system consistently and correctly identified the lock's status. Overall, this project delivered an efficient and low-cost security system that is easy to control, offering a practical new option for today's IoT-based smart homes.

Leveraging Ensemble Learning Technique for Efficient Fertilizer Recommendation

2026-01-02T08:56:20+00:00

Agricultural productivity plays a critical role in improving crop yields, and the suitable use of fertilizers plays a significant role in enhancing crop yield. Traditional fertilizer recommendation approaches often rely on generalized strategies that may not account for discrepancies in soil properties, climatic conditions. To address this limitation, we proposed an intelligent Fertilizer Recommendation System (FRS) using an Ensemble Learning method. This system integrates multiple ensemble learning models, such as Bagging, AdaBoost, GBoosting, Extra Trees, and CatBoost to enhance recommendation accuracy. The ensemble model is trained on soil parameters (N) nitrogen (P) phosphorus, (K) potassium, and moisture to recommend the optimal fertilizer type in Andhra Pradesh region, India. The result shows that all ensemble models utilized were effective, and CatBoost model has achieved 94.78% with highest accuracy, when compared with the other ensemble models.

Implementasi Prototipe SIEM Berbasis Wazuh pada Website dengan Pengujian FIM dan Threat Hunting

2026-01-02T08:54:48+00:00

This study implements a Wazuh-based Security Information and Event Management (SIEM) prototype to enhance security monitoring for a web application. The architecture uses two VPS instances: a web server as the log source equipped with a Wazuh Agent, and a monitoring server running the Wazuh Manager and Dashboard for event analysis and visualization. The evaluation combines threat hunting and File Integrity Monitoring (FIM) using several test scenarios: OWASP ZAP scanning, XSS, SQL injection (login-form testing and automated sqlmap attacks), and SSH brute force using hydra. The results show that Wazuh successfully detects XSS via rule 31105 (level 6) and sqlmap-based SQL injection via rule 31106 (level 6) because the attack patterns are clearly recorded in the web access logs. SSH brute force is strongly detected by rule 5763 (level 10), indicating repeated failed login attempts. In addition, FIM records file changes such as added and modified files (e.g., rules 554/550); however, it may generate noise when monitoring dynamic directories. The SQL injection attempt through the login form does not produce a specific SQL injection alert, suggesting limitations in log visibility/format and the need for decoder/ruleset tuning. Overall, Wazuh is effective for log-based security monitoring, while detection quality depends on log completeness, rule configuration, and FIM scope.

IoT-Based Prepaid Water Meter System with Digital Token for PDAM Customers

2026-01-02T08:55:57+00:00

Conventional PDAM water meter systems, which often employ post-paid techniques, frequently lead to payment delinquencies, challenges in consumption regulation, and delayed leak detection. This study creates a prototype of an IoT-enabled prepaid water meter system utilizing the ESP32 microcontroller, YF-S201 flow meter sensor, TDS sensor for water quality assessment, and a relay-operated solenoid valve, combined with a PHP/MySQL website hosted on AWS Academy for digital token transactions and remote monitoring. The technique employs a Research and Development (R&D) framework including analysis, design, implementation, and testing phases. The test findings indicate that the system efficiently conducts real-time monitoring of water consumption and quality while successfully addressing payment arrears; nevertheless, a significant constraint is present in the solenoid valve, which produces an insufficient water flow, failing to satisfy typical home flow rate requirements. The prototype has significant promise for updating PDAM water distribution by offering transparency and comprehensive control to consumers; however, enhancements to the solenoid valve components are necessary before practical adoption

BBRI Stock Trend and Volatility Analysis Using Moving Average and Standard Deviation

2026-01-02T08:55:38+00:00

Bank Rakyat Indonesia (BBRI) is one of the largest banks in Indonesia with significant market capitalization. This study aims to analyze price trend movements and stock volatility of BBRI for the period of January 2023 - October 2025 using Simple Moving Average (SMA) 20 and 50 days and Standard Deviation 20 days. Daily data (672 observations) were analyzed to identify trend patterns and investment risk levels. The results show that BBRI stock experienced a strong uptrend phase in 2023 (+33.1%), followed by a rally to a peak of IDR 5,763 in May 2024. However, the mid-2024 period was marked by high volatility with a whipsaw pattern (false signals), where the October 2024 Death Cross confirmed a final decline to the lowest level of IDR 3,176 (-44.9% from the peak). Volatility varied from 40 IDR (lowest, July-August 2023) to 425 IDR (highest, May-June 2024), indicating a negative correlation with trend strength. Moving Average is effective in identifying major trend changes, providing practical implications for technical analysis-based investment strategies.

The Effect of Telecommunication Services in Enhancing Economic Development in Banadir Region, Mogadishu, Somalia

2026-01-02T08:57:58+00:00

This study investigates the impact of telecommunication services provided by Hormuud Telecom on economic development in the Banadir region of Mogadishu, Somalia. Employing a quantitative approach, this study examines how broadband Internet penetration, network coverage, and Internet usage influence key economic indicators, such as GDP growth, employment, and financial inclusion. Primary data were collected through closed-ended questionnaires administered to a sample of 100 participants, including Hormuud Telecom department leaders, employees, and customers. Secondary data on company and economic metrics were also utilized. The collected data were analyzed using descriptive Statistics and analysis performed in SPSS to explore the relationship between telecommunications development and economic growth. The findings reveal a strong positive perception of the impact of Hormuud Telecom's services on macroeconomic development, particularly in terms of rising GDP per capita and the proliferation of mobile money services such as EVC Plus. However, the study also uncovers a lack of consensus regarding the micro-economic benefits, with polarized views on the impact on local business growth, unemployment rates, and poverty reduction issues. Furthermore, the results highlight a significant gap between network coverage expansion and the quality and reliability of Internet services. The study concludes that while telecommunications infrastructure is a crucial enabler of economic development in the Banadir region, its full potential is moderated by factors such as service quality, digital literacy, and broader socioeconomic challenges. The findings offer valuable insights for policymakers, regulators, and industry stakeholders in Somalia, emphasizing the need for more inclusive and sustainable strategies to ensure that the benefits of digital growth are widely shared.

Penerapan Teknologi Spark AR dalam Pengenalan Suntiang Minangkabau

2026-01-02T08:54:19+00:00

Penelitian ini menjelaskan penerapan teknologi Spark AR dalam pengenalan Suntiang Minangkabau. Suntiang Minangkabau adalah simbol budaya dan identitas Minangkabau yang biasanya selalu dipakai oleh pengantin wanita atau anak daro di daerah tersebut. Namun pengetahuan dan pemahaman mengenai suntiang ini masih terbatas, terutama pada kalangan generasi muda termasuk masyarakat Minangkabau itu sendiri. Oleh karena itu, penelitian ini memanfaatkan teknologi Augmented Reality(AR) menggunakan Spark AR untuk menghadirkan pengalaman digital interaktif yang memungkinkan pengguna untuk secara virtual mengenakan dan mengenal Suntiang Minangkabau melalui suatu platform seperti instagram. Metode perancangan filter ini melibatkan tahap pengumpulan data, perancangan, implementasi dan pengujian. Rancangan filternya juga dibuat dengan metode prototype dengan menampilkan gambar suntiang dalam bentuk 2D serta informasi tentang Suntiang Minangkabau tersebut. Hasil penelitian ini tentunya bertujuan bahwa penerapan teknologi Spark AR dapat menjadi alat yang efektif dalam menambah pengetahuan dan pembelajaran tentang budaya Minangkabau salah satunya tentang Suntiang Minangkabau. Selain itu juga bisa berpotensi untuk mempromosikan warisan budaya dan edukasi melalui media digital

Integration of Information Technology in School Leadership: Challenges and Innovations in Teacher Professional Development in Azerbaijan

2026-01-02T08:53:56+00:00

This study is to identify and analyze the difficulties that school principals face in supporting teachers' professional development. The subject of the study is the role of school principals in the professional development process and the obstacles they face. Qualitative methods - interviews and analysis - were used in the study. Analysis of the collected data shows that the main difficulties of principals are lack of time, resource and financial constraints, teacher motivation problems, lack of continuity of training, and weakness of pedagogical leadership skills. These factors hinder the effective organization and implementation of professional development. The results confirm the importance of strengthening the pedagogical leadership functions of school management, increasing resources, and creating continuous training systems. Also, improving the quality of teachers' professional development requires the development of school culture and motivational mechanisms. However, the findings also indicate a significant digital transformation through the establishment of educational information systems, e-learning platforms, and international cooperation to enhance teachers' digital competence. The study makes theoretical and practical contributions to improving teachers' professional development in the Azerbaijani education system and identifies key directions for future research

Klasifikasi Jenis Buah Mangga Menggunakan Convolutional Neural Network (CNN) Berbasis Citra Digital

2026-01-02T08:57:39+00:00

Sektor pertanian dan perdagangan Indonesia sangat dibantu oleh klasifikasi jenis mangga. Studi ini menerapkan Convolutional Neural Network (CNN) untuk mengklasifikasikan sembilan jenis mangga menggunakan gambar digital. Dataset dikumpulkan secara langsung melalui kamera smartphone, yang menghasilkan gambar awal untuk setiap kelas mangga. Sebelum proses pra-pengolahan, resize dan normalisasi dilakukan, dan model CNN dilatih untuk mengekstraksi fitur dan klasifikasi dengan menggunakan lapisan konvolusi dan dense. Hasil evaluasi menunjukkan bahwa model memiliki akurasi sekitar 90% dalam mengidentifikasi jenis mangga. Di masa depan, sistem ini diharapkan dapat membantu perkembangan pertanian cerdas dan membantu klasifikasi buah secara otomatis.

Analisis Aksesibilitas Tokopedia Berbasis Mobile Menggunakan User Experience Questionnaire

2026-01-02T08:55:20+00:00

This study aims to comprehensively examine the accessibility level and user experience of Tokopedia on mobile based on the Human-Computer Interaction (HCI) approach by employing the User Experience Questionnaire (UEQ) as the primary instrument. Data collection involved 50 active Tokopedia users, with the UEQ instrument comprising 26 statements and a 7-point response scale. Six core scales—attractiveness, perspicuity, efficiency, dependability, stimulation, and novelty—were evaluated and analyzed by calculating the average score for each aspect and comparing it to the standard UEQ benchmark. The results demonstrate that Tokopedia excels in attractiveness, perspicuity, and efficiency, while innovation in the novelty and stimulation scales remains a challenge that requires further optimization. This study underlines the importance of UX evaluation based on HCI as a foundation for developing inclusive features to support an optimal and sustainable user experience