cover
Article Per Year (5 Year)
All
Home Page
OAI Link
Editorial Team
Contact
Reviewer
Google Scholar
Contact Name
M. Miftach Fakhri
Contact Email
fakhri@unm.ac.id
Phone
+6282290603030
Journal Mail Official
wahid@unm.ac.id
Editorial Address
Program Studi Teknik Komputer, UNM Parangtambung, Daeng Tata Raya, Makassar, South Sulawesi, Indonesia
Location
Kota makassar,
Sulawesi selatan
INDONESIA
Journal of Embedded Systems, Security and Intelligent Systems
Published by Universitas Negeri Makassar
ISSN : 2745925X     EISSN : 2722273X     DOI : -
Core Subject : Science,
Computer Science & IT
The Journal of Embedded System Security and Intelligent System (JESSI), ISSN/e-ISSN 2745-925X/2722-273X covers all topics of technology in the field of embedded system, computer and network security, and intelligence system as well as innovative and productive ideas related to emerging technology and computer engineering, including but not limited to : Network Security System Security Information Security Social Network & Digital Security Cyber Crime Machine Learning Decision Support System Intelligent System Fuzzy System Evolutionary Computating Internet of Thing Micro & Nano Technology Sensor Network Renewable Energy Wearable Devices Embedded Robotics Microcontroller
Arjuna Subject : Ilmu Komputer (semua kategori) - Kecerdasan Buatan
Articles 157 Documents
 12   13   14   15   16 
Fire Detection and Room Firefighting System Based on IoT Using C4.5 Decision Tree Algorithm Ismayanti, Rika; Triadi, Fara; Jaya, Arsan Kumala; Irawan, Ade
Journal of Embedded Systems, Security and Intelligent Systems Vol 6, No 4 (2025): Desember 2025
Publisher : Program Studi Teknik Komputer

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.59562/jessi.v6i4.10685

Abstract

Early fire detection is a critical requirement in indoor safety systems, where delays of only a few seconds can escalate into severe damage and casualties. Conventional devices often rely on single-sensor thresholds, which are highly susceptible to false alarms and unstable performance in dynamic indoor environments. This study develops an Internet of Things (IoT)-based multi-sensor fire detection and autonomous firefighting system integrated with a C4.5 decision tree classifier for real-time hazard recognition and short-term risk prediction. The prototype combines DHT22 temperature, MQ-135 gas, infrared flame, and ultrasonic water-level sensors with an ESP32 microcontroller, servo-controlled nozzle, and pump-based water spraying, all connected to an Android–Firebase platform for remote monitoring. A multivariate time-series dataset of 200 sensor sequences was preprocessed using a five-step sliding-window model and evaluated through 1,000 repeated hold-out trials. The C4.5 classifier achieved a mean accuracy of 84.9%, with peak values exceeding 90%, and clearly separated Safe, Alert, and Danger states, with smoke concentration emerging as the dominant predictor. Experimental tests in a 60 × 40 × 30 cm chamber produced 1–2 s reaction times, eight successful extinguishing events, and four failures attributable to mechanical belt detachment rather than model errors. These findings indicate that interpretable decision-tree models, when combined with IoT sensing and autonomous actuation, can provide a low-cost framework for real-time fire warning and automatic suppression. Future work should address mechanical robustness, extended deployment, and multi-room scalability
Performance Comparison of Svm and Naïve Bayes For Indonesian-Language Sentiment Analysis On Free Fire Reviews Using Tf-Idf And Smote Wahid, Yokogeri Abdullah; Sanatang; Andayani, Dyah Darma
Journal of Embedded Systems, Security and Intelligent Systems Vol 6, No 4 (2025): Desember 2025
Publisher : Program Studi Teknik Komputer

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.59562/jessi.v6i4.10818

Abstract

The popularity of online games continues to increase, including Free Fire, which has gained more than one billion downloads and millions of user reviews on the Google Play Store. However, the variation and inconsistency of user comments make manual sentiment evaluation difficult. This study aims to compare the performance of Support Vector Machine (SVM) and Naïve Bayes in classifying user review sentiment on the Free Fire game. A total of 535 Indonesian-language reviews were collected using web scraping and processed through text cleaning, case folding, normalization, stopword removal, and stemming. Sentiment labels were assigned manually based on review content. The dataset was divided into training and testing using a 70:30 ratio, and feature extraction used Term Frequency–Inverse Document Frequency (TF-IDF). Two scenarios were implemented: a baseline without class balancing and a scenario using Synthetic Minority Oversampling Technique (SMOTE) to address class imbalance. Results show that SVM outperforms Naïve Bayes in both scenarios. In the baseline, SVM achieved 89.81% accuracy, while Naïve Bayes obtained 82.80%. After SMOTE, SVM improved to 91.08% accuracy and Naïve Bayes to 89.17%. These findings indicate that SVM, especially with SMOTE, provides a more effective and balanced performance for sentiment classification on Free Fire reviews. The study contributes to providing a more accurate understanding of user perception and strengthening model development for sentiment analysis on digital game applications.
Energy Audit Integrated with Fuzzy Neural Network Predictive Maintenance for Central Chillers Saragih, Budiman R; Aldi Cahya Muhammad
Journal of Embedded Systems, Security and Intelligent Systems Vol 6, No 4 (2025): Desember 2025
Publisher : Program Studi Teknik Komputer

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.59562/jessi.v6i4.10857

Abstract

Because central chiller systems significantly affect electricity usage in office buildings, predictive maintenance and energy audits would be important to increase efficiency. This research analyzes the data from the thermodynamic audit and the central chillers and the monthly electricity usage to assess the energy performance of the East Jakarta Mayor's Office Building A for the years 2023-2024. Based on the Building A, in 2024, the total estimated electrical consumption will be 2,019,550 kWh. This results in total energy use intensity of 106.9 kWh/m²/year. Based on the estimated data, the HVAC systems use more than half of the total electrical consumption. The simulations show, for the data provided, the energy efficiency measures have a saving potential of approximately 728,847 kWh/year which equals 36.1% on a total consumption of 2,019,550 which would also save 36.2% of 1.26 billion/year. The total energy use intensity would be reduced to 90.5 kWh/m²/year, with the emission reduction of approximately 604.9 tCO₂e/year. Based on the consumed data and the paired t test on the 12 sampled data the results would show, with p value < 0.001, a 97,438 kWh/month average reduction in electrical consumption in the 2023-2024 years, which shows a correlation in the expected data with operational and standard fix measures. The Fuzzy Neural Network is, and can be used with other data to show other measures of predictive maintenance rather than the conventional audit based measures used.
Implementation of Automated Intelligent Irrigation and Fertilization System Based on the Internet of Things for Home Wine Hobbyists Parenreng, Jumadi M.; Lamada, Mustari S; Wahid, M. Syahid Nur; Wahid, Abdul; Azra, Muh. Azfa
Journal of Embedded Systems, Security and Intelligent Systems Vol 6, No 4 (2025): Desember 2025
Publisher : Program Studi Teknik Komputer

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.59562/jessi.v6i4.11207

Abstract

The agricultural sector faces serious challenges due to climate change and rising global food needs, which demand more efficient implementation of water and plant nutrient management. Viticulture, particularly on an urban scale, requires precise irrigation and fertilization arrangements because grapes are very sensitive to water availability and nutrient concentrations at certain growth phases. This research aims to design and implement a prototype of an Internet of Things (IoT)-based Smart Nutrition Irrigation System for home wine hobbyists using the Research and Development (R&D) method. The developed system is able to monitor environmental and soil conditions in real-time and control irrigation and fertilization automatically based on predetermined parameters. The results of the performance evaluation showed that the system had an average response time of ±1–3 seconds from the time the sensor data was received until the actuator was activated. TDS sensors are able to detect changes in the concentration of fertilizer solutions in the range of 0–1200 ppm, thus supporting the quality control of nutrient solutions. The results of functional tests show that the system successfully activates irrigation automatically when soil moisture is below the 75% threshold and runs scheduled fertilization consistently. In addition, the system is able to send real-time monitoring and actuator status notifications through an online application with a 100% message success rate during testing. Overall, the system has been shown to improve the efficiency of water and nutrient management in urban grape cultivation and has the potential to be further developed on a larger scale and applied to precision agriculture
A Comparative Study of Decision Tree and Gradient Boosting Tree Algorithms for Predicting College Enrollment Decisions of High School Students Aras, Rezty Amalia; Utami Kusuma Dewi; Yabes Dwi Nugroho
Journal of Embedded Systems, Security and Intelligent Systems Vol 7 No 1 (2026): March 2026
Publisher : Program Studi Teknik Komputer

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.59562/jessi.v7i1.2601

Abstract

The declining interest of high school students in pursuing higher education has become a major concern in Indonesia's education sector. This study aims to develop a data-driven predictive model to assist schools in identifying students’ decisions regarding further education. The study compares two popular classification algorithms, Decision Tree and Gradient Boosted Tree, using a dataset of 300 high school students comprising 10 attributes such as school accreditation, parental income, interest level, and residential status. The research method involves data preprocessing, model training, and performance evaluation using a confusion matrix to measure accuracy, precision, and recall. The results show that the Decision Tree algorithm achieved an accuracy of 76.67%, with a precision of 78.57% and a recall of 73.33% for the "college" class. Meanwhile, the Gradient Boosted Tree produced an accuracy of 73.33%, with a strength in recall for the "not attending college" class at 80%, but was less optimal in detecting students who pursued higher education. It can be concluded that the Decision Tree outperforms in terms of accuracy and interpretability, making it more suitable for use in school environments as a decision-support tool for early intervention, scholarship programs, and career counseling.
Ensemble-Based Clickbait Detection in Indonesian Online News Surianto, Dewi Fatmawati; Diny Anggriani Adnas
Journal of Embedded Systems, Security and Intelligent Systems Vol 6, No 4 (2025): Desember 2025
Publisher : Program Studi Teknik Komputer

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.59562/jessi.v6i4.11251

Abstract

Clickbait has become a pervasive issue in online news media, particularly in the Indonesian digital information ecosystem, where sensational headlines are frequently used to attract user attention at the expense of content accuracy. This phenomenon not only degrades information quality but also contributes to the spread of misinformation. To address this challenge, this study proposes an ensemble-based machine learning approach for detecting clickbait in Indonesian-language news articles by jointly analyzing headlines and full article content. The proposed method employs Term Frequency–Inverse Document Frequency (TF-IDF) feature extraction with extended n-gram configurations to capture both lexical and contextual patterns characteristic of clickbait. Three baseline classifiers, Multinomial Naïve Bayes, Logistic Regression, and Support Vector Machine are integrated into a hard voting ensemble framework to leverage their complementary strengths. The experiments were conducted on the CLICK-ID dataset, consisting of annotated Indonesian news articles, using an 80:20 train–test split. Experimental results demonstrate that the proposed ensemble model outperforms all individual baseline classifiers, achieving an overall accuracy and F1-score of 93%. The ensemble approach shows notable improvements in recall for the clickbait class, indicating its effectiveness in minimizing false negatives. Furthermore, qualitative analysis using word cloud and bigram visualization reveals distinct linguistic patterns between clickbait and non-clickbait articles, supporting the discriminative capability of the extracted features. These findings confirm that combining TF-IDF with ensemble learning provides a robust and effective solution for clickbait detection in Indonesian online news. The proposed model contributes to the development of more reliable content filtering systems and supports efforts to improve information quality in digital media environments.
Performance Evaluation of Fiber Optic versus Copper Cable Networks: A Comparative Study Muhammad Bitrayoga; Evan Haviana; Jeffrey Payung Langi; Sulastri Kakaly; Apriani Riyanti
Journal of Embedded Systems, Security and Intelligent Systems Vol 6, No 4 (2025): Desember 2025
Publisher : Program Studi Teknik Komputer

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.59562/jessi.v6i4.11443

Abstract

This study presents a comparative performance evaluation of Fiber Ethernet and copper Ethernet (Cat6) using a controlled testbed to quantify differences in throughput and Quality of Service (QoS). Identical endpoints are connected through a managed switch, and the physical medium is alternated between a fiber link and a Cat6 link. Traffic is generated under TCP and UDP at three utilization levels (10%, 50%, and 90%) and three distance conditions (10 m, 50 m, and 90–100 m for Cat6). The evaluation uses application-level and interface-level measurements, including TCP/UDP throughput, round-trip latency, jitter, packet loss, and reliability counters such as CRC/FCS errors and TCP retransmissions. Results show that both media can achieve near line-rate throughput at short distance and low load, but fiber exhibits consistently lower latency and jitter and near-zero loss across conditions. As utilization increases, Cat6 displays larger variability in delay and a higher tendency toward UDP loss, indicating reduced timing stability under stress. At longer copper runs (90–100 m), the gap widens: Cat6 shows higher jitter and loss and increased error-related counters, while fiber remains stable. These findings suggest that the main advantage of fiber is not only peak capacity but also QoS predictability and link integrity, which are critical for real-time and high-utilization services. The study provides practical guidance for selecting media in campus and building networks where distance, load, and service requirements must be balanced. Future work will include electromagnetic-noise trials, different switch/NIC models, and statistical tests to validate effect sizes across repetitions in diverse building environments
Comparative Performance Analysis of Web Caching and Load Balancing in Web-Based E-Learning Platforms Silamba, Yurico Ignasius May; Utomo, Muhammad Nur Yasir; Syamsuddin, Irfan
Journal of Embedded Systems, Security and Intelligent Systems Vol 7 No 1 (2026): March 2026
Publisher : Program Studi Teknik Komputer

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.59562/jessi.v7i1.2602

Abstract

Purpose – The increasing adoption of web-based E-learning platforms has introduced significant performance challenges under high user load, including elevated latency, degraded response time, and service downtime. While web caching and load balancing are commonly employed as mitigation strategies, empirical guidance on which technique better suits E-learning traffic characteristics remains limited. This study aims to provide a direct, controlled comparison between the two approaches.Design/methods/approach – An experimental methodology was employed using a Moodle-based E-learning platform deployed in a controlled virtualized environment.  Web caching was implemented with Varnish Cache and load balancing with HAProxy, each deployed on separate virtual machines running Moodle as the E-learning platform. Apache JMeter was used to simulate concurrent workloads ranging from 50 to 250 users, measuring latency, throughput, server response time, and failure rate.Findings – The results show that web caching consistently outperformed load balancing across all metrics. At 250 concurrent users, web caching reduced latency by 9.9%, improved throughput by approximately 4.5 times, decreased server response time by 96.6%, and lowered failure rate by 80.2% compared to load balancing. These findings indicate that caching more effectively mitigates backend overload in E-learning systems dominated by repetitive content access.Research implications/limitations – The experiments were conducted in a virtualized environment and focused primarily on static and semi-static content. Hybrid architecture combining both techniques were not evaluated.Originality/value – This study provides a head-to-head empirical comparison between web caching and load balancing under identical conditions, offering practical architectural guidance for infrastructure planning in academic environments.
The Effect of K-NN Algorithm Practice Chatbot Tutors on the Speed and Accuracy of Solving AI Problems in Vocational Schools Andi Asfar; Andi Winda Purnamasari; Angriani Nur; Aprilia; Arnis; Ashabul Kahfi
Journal of Embedded Systems, Security and Intelligent Systems Vol 6, No 1 (2025): March 2025
Publisher : Program Studi Teknik Komputer

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.59562/wrd2bn63

Abstract

Learning Artificial Intelligence (AI) concepts at the vocational high school level often presents challenges, particularly in understanding procedural algorithms such as K-Nearest Neighbor (K-NN). Students frequently experience difficulties in calculating distances, selecting the optimal value of k, and determining classification results accurately and efficiently. This study investigates the effect of a chatbot tutor designed for K-NN practice on students’ speed and accuracy in solving AI problems. A quasi-experimental design with a pretest–posttest control group model was employed. Sixty students from a vocational high school majoring in Software Engineering were divided into an experimental group receiving chatbot-assisted practice and a control group receiving conventional instruction. Data were collected through validated K-NN problem-solving tests and task completion time measurements. Statistical analysis using paired and independent sample t-tests revealed that the experimental group demonstrated significantly faster completion times and higher accuracy scores compared to the control group (p < 0.05). The effect size indicated a moderate to high practical impact. The findings suggest that chatbot-based tutoring can enhance both efficiency and precision in learning K-NN algorithms. This study contributes empirical evidence supporting the integration of chatbot tutors into Artificial Intelligence instruction in vocational education to improve problem-solving performance.
Classification of Students' Emotions from Facial Expressions Using CNN to Support Adaptive Learning Akmal Hidayat; Hera Ariska; Iren Kirana; Asmiyah Auliatna; Dian Sri Yuninda; Elvira Nur
Journal of Embedded Systems, Security and Intelligent Systems Vol 6, No 1 (2025): March 2025
Publisher : Program Studi Teknik Komputer

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.59562/b1rcm003

Abstract

The integration of affective aspects into adaptive learning systems remains limited, as most educational technologies primarily rely on cognitive performance indicators. However, students’ emotional conditions significantly influence engagement, motivation, and learning outcomes. This study aims to develop and evaluate a Convolutional Neural Network (CNN) model for classifying students’ emotions based on facial expressions to support adaptive learning environments. A quantitative experimental approach was employed. Facial expression image data were preprocessed through face detection, resizing, normalization, and data augmentation before being trained using a CNN architecture with the Adam optimizer and categorical cross-entropy loss function. Model performance was evaluated using accuracy, precision, recall, F1-score, and confusion matrix analysis. The experimental results show that the proposed CNN model achieved an overall accuracy of 90% with an average F1-score of 0.88 across four emotion categories (Happy, Sad, Neutral, and Angry). The confusion matrix indicates that most predictions fall within the correct classification range, although minor misclassifications occurred between low-intensity Sad and Neutral expressions. The stability of training and validation loss curves demonstrates good generalization ability without significant overfitting. These findings indicate that CNN-based facial emotion classification can serve as a reliable component in adaptive learning systems by providing real-time affective feedback. The study contributes to the development of artificial intelligence applications in education by integrating emotional recognition into adaptive instructional design

Page 14 of 16 | Total Record : 157

 12   13   14   15   16 

Filter by Year

2020 2026


Reset
Filter By Issues
All Issue Vol 7 No 1 (2026): March 2026 Vol 6, No 4 (2025): Desember 2025 Vol 6, No 3 (2025): September 2025 Vol 6, No 2 (2025): June 2025 Vol 6, No 1 (2025): March 2025 Vol 5, No 3 (2024): November 2024 Vol 5, No 2 (2024): July 2024 Vol 5, No 1 (2024): March 2024 Vol. 4 No. 1 (2023): Vol 4, No 1 (2023): May 2023 Vol. 3 No. 2 (2022): Vol 3, No 2 (2022): November 2022 Vol. 3 No. 1 (2022): Vol 3, No 1 (2022): May 2022 Vol. 2 No. 2 (2021): Vol 2, No 2 (2021): November 2021 Vol. 2 No. 1 (2021): Vol 2, No 1 (2021): May 2021 Vol. 1 No. 2 (2020): Vol 1, No 2 (2020): November 2020 Vol. 1 No. 1 (2020): Vol 1, No 1 (2020): May 2020 More Issue