cover
Contact Name
Hendra Nelva Saputra
Contact Email
hendra.nelva@umkendari.ac.id
Phone
+6282193165892
Journal Mail Official
hendra.nelva@umkendari.ac.id
Editorial Address
Jl. K.H. Ahmad Dahlan No. 10, Gedung E Lantai 2
Location
Kota kendari,
Sulawesi tenggara
INDONESIA
Decode: Jurnal Pendidikan Teknologi Informasi
ISSN : 27752984     EISSN : 27751813     DOI : https://doi.org/10.51454/decode
DECODE: Jurnal Pendidikan Teknologi Informasi publishes articles in English and Indonesian which will be published 3 times a year, namely March, July and November. Articles that can be considered for publication in this journal are the results of research that is in accordance with the focus and scope of the journal including: (1) Software Engineering, (2) Data Mining, (3) Artificial Intelligence, (4) Computer Network, (5) Robotics System, (6) Information Security, (7) Mobile Learning, (8) Online Learning, (9) Multimedia, (10) Development Learning Media, (11) Integration of Technology in Learning.
Articles 342 Documents
Penerapan Inception V3 Untuk Deteksi Penyakit Tuberkulosis Berdasarkan Citra Rontgen Dada Melalui Sistem Berbasis Web Davit Saputra Purba; Juliansyah Putra Tanjung; Juanda Putra Sinaga; Febrianta Tarigan; Mhd Allifi Rifqi
Decode: Jurnal Pendidikan Teknologi Informasi Vol. 5 No. 3: NOVEMBER 2025
Publisher : Program Studi Pendidikan Teknologi Infromasi UMK

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.51454/decode.v5i3.1228

Abstract

Tuberkulosis (TB) merupakan penyakit menular yang masih menjadi masalah serius di sektor kesehatan, khususnya di Indonesia. Proses diagnosis melalui citra rontgen dada sering mengalami kendala karena keterbatasan jumlah radiolog dan kemiripan visual antara TB dan gangguan paru lainnya. Penelitian ini menerapkan model deep learning dengan arsitektur Inception V3 untuk mendeteksi TB berdasarkan citra rontgen dada. Model dikembangkan dalam sistem berbasis web menggunakan framework Django, sehingga dapat digunakan secara interaktif oleh pengguna. Dataset diambil dari platform Kaggle, terdiri atas 4.200 citra yang terbagi dalam dua kelas, yaitu Normal dan Tuberculosis. Proses fine-tuning dilakukan dengan 30 epoch, learning rate sebesar 0,0001, dan dropout rate 0,5. Model diuji menggunakan confusion matrix untuk menilai performa klasifikasinya. Hasil pelatihan menunjukkan akurasi sebesar 97% dengan loss 0,08, sedangkan pengujian menghasilkan akurasi 98,57%. Nilai precision, recall, dan f1-score rata-rata mencapai 0,99, menunjukkan kemampuan model yang sangat baik dalam membedakan citra paru normal dan terinfeksi TB. Sistem ini diharapkan dapat membantu proses diagnosis awal secara cepat dan akurat, serta menjadi dasar pengembangan teknologi deteksi penyakit berbasis pembelajaran mendalam di bidang medis.
Model Convolutional Neural Network (CNN) Custom Sequential untuk Klasifikasi Citra Ikan Hias Carassius Auratus pada Industri Akuakultur Rodiah; Diana Tri Susetianingtias; Eka Patriya
Decode: Jurnal Pendidikan Teknologi Informasi Vol. 5 No. 3: NOVEMBER 2025
Publisher : Program Studi Pendidikan Teknologi Infromasi UMK

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.51454/decode.v5i3.1229

Abstract

Identifikasi varietas ikan hias Carassius auratus dengan keanekaragaman morfologi yang tinggi masih sering dilakukan secara manual, sehingga menimbulkan kesulitan bagi para pecinta ikan hias maupun pelaku akuakultur dalam membedakan jenisnya secara tepat. Kondisi ini menuntut adanya solusi berbasis teknologi yang mampu memberikan hasil identifikasi lebih cepat, praktis, dan presisi. Penelitian ini bertujuan untuk mengembangkan model klasifikasi citra menggunakan pendekatan Convolutional Neural Network (CNN) dengan arsitektur custom sequential sebagai sistem identifikasi visual otomatis. Tahapan penelitian dilakukan melalui pengumpulan 2.080 citra ikan yang mewakili delapan kelas varietas, dilanjutkan dengan proses pra-pemrosesan data agar citra siap digunakan pada tahap pelatihan model. Arsitektur CNN yang dibangun terdiri dari tujuh lapisan konvolusi dengan jumlah filter bertingkat mulai dari 32 hingga 256, disertai proses pooling, aktivasi ReLU, transformasi flatten, serta lapisan fully connected dengan mekanisme dropout untuk mengurangi overfitting, dan diakhiri dengan fungsi aktivasi softmax sebagai klasifikasi multi-kelas. Hasil eksperimen menunjukkan bahwa model mampu mengklasifikasikan citra ikan hias dengan tingkat akurasi mencapai 93,03% serta menghasilkan prediksi data uji secara konsisten tepat. Penelitian ini berkontribusi dalam menghadirkan pendekatan klasifikasi berbasis citra yang efisien, akurat, dan aplikatif, sehingga dapat mendukung proses identifikasi varietas ikan hias dalam industri akuakultur secara lebih modern dan berbasis teknologi.
Sentiment Analysis of User Reviews on Maxim Application Using the Long Short-Term Memory (LSTM) Methods Maria Ilona Junide Bria; Anik Vega Vitianingsih; Anastasia Lidya Maukar; SY. Yuliani; Pamudi Pamudi
Decode: Jurnal Pendidikan Teknologi Informasi Vol. 5 No. 3: NOVEMBER 2025
Publisher : Program Studi Pendidikan Teknologi Infromasi UMK

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.51454/decode.v5i3.1257

Abstract

The technological developments have encouraged the emergence of app-based transportation services that are increasingly popular with the public, one of which is the Maxim app. Despite offering convenience in booking transportation and other services, this app still receives various reviews from users regarding service quality. User feedback is provided through the Maxim app review section available on the Google Play Store platform. Sentiment analysis is applied in this study to identify shortcomings in the Maxim app, to help developers improve service quality and understand user satisfaction. The research procedure it comprises several phases, including data collection, text preprocessing, determining sentiment labels, assigning weights to terms, and a classification process using the Long Short-Term Memory (LSTM) algorithm. This studi unlike previous studies that commonly used classical machine learning techniques including Naïve Bayes and SVM, or BiLSTM, this research applies an LSTM model with lexicon-based sentiment labeling to improve consistency and contextual understanding in sentiment classification. A confusion matrix was utilized to evaluate the model’s performance. Overall, 1,200 user reviews were gathered through web scraping techniques from June 2024 to June 2025. The sentiment classification process uses a lexicon-based method to categorize user reviews grouped into three sentiment classes: positive, neutral, and negative. The findings suggest that 762 reviews are labelled as positive, 157 as neutral, and 281 as negative. The LSTM method testing demonstrated excellent performance, achieved 95.21% accuracy, 97.22% precision, 84.02% recall, and an F1-score of 88.84%.
Implementasi Metode Fuzzy Tsukamoto pada Alat Tetas Telur untuk Monitoring dan Controlling Suhu dan Kelembapan Aghnia Nurhidayah; Anandiaz Agung Pradana; Muhammad Rifan Fathoni; Anindita Septiarini; Masna Wati
Decode: Jurnal Pendidikan Teknologi Informasi Vol. 5 No. 3: NOVEMBER 2025
Publisher : Program Studi Pendidikan Teknologi Infromasi UMK

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.51454/decode.v5i3.1258

Abstract

Inkubator atau alat tetas telur berfungsi menciptakan kondisi lingkungan buatan yang menyerupai pengeraman alami, khususnya dalam menjaga suhu dan kelembapan. Ketidakstabilan dua parameter ini sering kali menjadi penyebab rendahnya tingkat keberhasilan penetasan telur ayam. Penelitian ini bertujuan untuk merancang dan mengimplementasikan sistem inkubator otomatis berbasis Internet of Things (IoT) yang dilengkapi dengan algoritma Fuzzy Tsukamoto. Sistem ini menggunakan NodeMCU sebagai mikrokontroler, sensor DHT11 untuk membaca suhu dan kelembapan, serta aplikasi Blynk sebagai antarmuka pemantauan. Sistem ini juga dirancang secara otomatis untuk mencatat data suhu dan kelembapan setiap 10 detik dan mengatur jumlah lampu pemanas yang menyala berdasarkan hasil inferensi fuzzy. Hasil pengujian menunjukkan bahwa sistem mampu menstabilkan suhu inkubator dalam waktu 33 menit pada kisaran 37°C–37,1°C dan kelembapan antara 55%–57%. Selain itu, sistem menunjukkan kemampuan adaptif terhadap perubahan suhu lingkungan dengan menyesuaikan jumlah lampu yang menyala. Sistem ini terbukti efektif dalam mengatur kondisi inkubasi secara otomatis dan akurat, serta mampu mengurangi ketergantungan terhadap pengaturan manual. Untuk pengembangan lebih lanjut, sistem ini dapat diintegrasikan dengan antarmuka berbasis mobile dan diuji pada skala yang lebih besar.
Smart Home Control Prototype Based on Internet of Things Using Google Assistant Muhammad Ridwan; Antoni Pribadi; Andri Nofiar Am
Decode: Jurnal Pendidikan Teknologi Informasi Vol. 5 No. 3: NOVEMBER 2025
Publisher : Program Studi Pendidikan Teknologi Infromasi UMK

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.51454/decode.v5i3.1274

Abstract

Smart home technology allows users to control household electronic devices automatically and remotely, including through voice commands. This study aims to design and develop a smart home control prototype based on the Internet of Things (IoT) using Google Assistant to control devices such as lights and fans. The system utilizes a NodeMCU ESP32 as the microcontroller, a 4-channel relay module as an automatic switch, and the Sinric Pro cloud service to connect voice commands  from  Google  Assistant  to  the  hardware.  The  research  follows  the ADDIE development model, which includes analysis, design, development, implementation, and evaluation stages. The test results show that the system can receive and execute voice commands with high accuracy in both Indonesian and English. However, the system still has limitations, such as dependence on a stable internet connection. Overall, this prototype functions well and has the potential to be further developed. In the Google Assistant analysis obtained from this system, Conversion and Calculation have the highest accuracy rate with 96.7%, while Voice command is the lowest with 85.3%.
Analisis Performa Sistem Informasi Akademik: Pengaruh Struktur Memori dan Manajemen Proses pada Sistem Operasi Linux Menggunakan Metode Eksperimental Kuantitatif Muhammad Rafli Ramadhan; Rinna Rachmatika
Decode: Jurnal Pendidikan Teknologi Informasi Vol. 5 No. 3: NOVEMBER 2025
Publisher : Program Studi Pendidikan Teknologi Infromasi UMK

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.51454/decode.v5i3.1272

Abstract

Penelitian ini menganalisis pengaruh struktur memori dan manajemen proses terhadap performa Sistem Informasi Akademik berbasis web pada sistem operasi Linux Ubuntu 22.04 LTS. Fokus utama penelitian adalah menilai efisiensi pengelolaan sumber daya sistem operasi dalam mendukung kinerja website akademik unpam.ac.id. Metode yang digunakan adalah eksperimen kuantitatif melalui pengujian langsung menggunakan alat ukur htop, vmstat, time, curl, dan GTmetrix. Hasil pengujian internal menunjukkan rata-rata 129 proses aktif dengan penggunaan memori sebesar 3,3 GB dari total 3,81 GB dan waktu akses rata-rata 2,45 detik, menandakan stabilitas sistem serta efisiensi manajemen proses dan memori. Sementara itu, pengujian eksternal menggunakan GTmetrix menunjukkan skor performa 56% (grade D), dengan Largest Contentful Paint (LCP) sebesar 7,7 detik dan Total Blocking Time (TBT) 66 ms. Hasil ini menunjukkan bahwa meskipun Linux mampu mengelola sumber daya secara efisien, performa website masih memerlukan optimalisasi di sisi front-end seperti penerapan CDN, kompresi file, dan HTTP/2. Penelitian ini memberikan kontribusi empiris terhadap pemahaman hubungan antara efisiensi sistem operasi, struktur memori, dan manajemen proses terhadap kinerja aplikasi web, serta dapat dijadikan acuan dalam pengembangan dan optimasi sistem informasi akademik di masa mendatang.
Improving Information Technology Service Management Effectiveness through ITIL Implementation at SMPN 7 IT Manggelewa Hardyanti; Yuyun Yuningsih RS; Maylisda Naomi Loko
Decode: Jurnal Pendidikan Teknologi Informasi Vol. 5 No. 3: NOVEMBER 2025
Publisher : Program Studi Pendidikan Teknologi Infromasi UMK

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.51454/decode.v5i3.1317

Abstract

Information Technology (IT) services are essential in enhancing the efficiency of administrative and learning processes in schools. However, many public junior high schools still manage IT services manually, without clear documentation or standard operating procedures, which leads to delayed responses, system unreliability, and low user satisfaction. This study aims to improve the effectiveness of IT service management at SMPN 7 IT Manggelewa through the implementation of a simplified model based on the ITIL (Information Technology Infrastructure Library) Service Operation framework. Employing a qualitative descriptive approach with an evaluative and model development orientation, data were collected through observations, interviews, document analysis, and questionnaires involving school stakeholders. The findings reveal major gaps between current practices and ITIL standards across five key processes: Incident Management, Request Fulfillment, Access Management, Problem Management, and Event Management. As a result, a contextualized ITIL-based model was developed, including SOPs, service request forms, access control procedures, a root-cause analysis template, and manual monitoring tools. The model was validated through a focus group discussion and revised for school-level applicability. This study concludes that the application of a school-adapted ITIL model improves service responsiveness, documentation practices, and stakeholder accountability in managing IT services.
Implementasi Deteksi Intrusi Aplikasi Web Berbasis Supervised Machine Learning: Studi Kasus LMS STT Terpadu Nurul Fikri Henry Saptono; Yuda Fatahillah Achmar; Hendro Sasongko Hadi; Salman Fathy Shiroth; Lambang Ramadhian Putra Aria; Muhamad Masayid Alfarizqi
Decode: Jurnal Pendidikan Teknologi Informasi Vol. 5 No. 3: NOVEMBER 2025
Publisher : Program Studi Pendidikan Teknologi Infromasi UMK

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.51454/decode.v5i3.1313

Abstract

Sistem deteksi intrusi merupakan komponen penting dalam menjaga keamanan aplikasi web, terutama pada platform pembelajaran daring seperti Learning   Management   System (LMS). Penelitian ini bertujuan mengimplementasikan sistem deteksi intrusi berbasis supervised machine learning untuk mengidentifikasi serangan SQL  Injection dan Cross-Site  Scripting (XSS) melalui analisis payload HTTP yang diterima oleh LMS Moodle. Model yang digunakan adalah algoritma Random Forest dengan representasi fitur berbasis (Term Frequency–Inverse Document Frequency) TF-IDF pada level karakter. Data pelatihan berasal dari gabungan dataset publik dan log aktivitas LMS internal yang telah melalui proses preprocessing serta masking data sensitif. Arsitektur sistem dirancang menggunakan plugin middleware pada LMS untuk menangkap log secara real-time, Redis sebagai message broker, dan Flask-RQ sebagai worker pemrosesan model, serta dashboard Grafana-Loki untuk visualisasi hasil deteksi. Hasil pengujian menunjukkan bahwa model Random Forest mencapai akurasi 99,94% dengan nilai presisi, recall, dan AUC yang sangat tinggi, menunjukkan kemampuan deteksi yang andal terhadap serangan SQL  Injection dan XSS. Sistem ini mampu beroperasi secara real-time tanpa mengganggu kinerja LMS, sehingga efektif diterapkan sebagai solusi keamanan siber pada lingkungan pendidikan. Implementasi ini berpotensi dikembangkan lebih lanjut untuk mendeteksi jenis serangan web lainnya secara adaptif.
Development of a Hybrid Voting Model with SMOTE and Random Search for Classification of Religious Facility Grant Recipients Muhammad Jamaris; Rita Novita
Decode: Jurnal Pendidikan Teknologi Informasi Vol. 5 No. 3: NOVEMBER 2025
Publisher : Program Studi Pendidikan Teknologi Infromasi UMK

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.51454/decode.v5i3.1342

Abstract

The process of determining recipients of religious facility grants requires high accuracy to ensure that aid is distributed fairly and supports equitable community services. Manual selection methods often face challenges such as data imbalance, diverse assessment criteria, and subjective decision-making, which can reduce accuracy and efficiency. This study proposes a hybrid machine learning model using Voting Ensemble (Hard and Soft), combining Logistic Regression (LR), Support Vector Machine (SVM), and K-Nearest Neighbor (KNN), optimized with Random Search and supported by SMOTE to handle class imbalance. The dataset consists of religious facility grant applications in Riau Province, with preprocessing, SMOTE balancing, and Stratified K-Fold Cross Validation applied for robust evaluation. The experimental results show that the Hybrid Voting model outperforms single models, achieving an average accuracy of 99.46%, with precision, recall, and F1-score consistently above 96%, and some folds achieving 100% accuracy. These findings demonstrate that the hybrid approach enhances prediction stability, reduces misclassification of minority classes, and provides a decision-support system that is objective, accurate, and efficient for grant recipient selection.
Integrating Staggered and Collaborative Learning for Ethical AI-Assisted Academic Writing Soetam Rizky Wicaksono; Rudy Setiawan; Muhammad Nurwegiono
Decode: Jurnal Pendidikan Teknologi Informasi Vol. 5 No. 3: NOVEMBER 2025
Publisher : Program Studi Pendidikan Teknologi Infromasi UMK

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.51454/decode.v5i3.1340

Abstract

The rapid advancement of artificial intelligence (AI) has prompted a paradigm shift in higher education, demanding innovative instructional models that not only leverage technological tools but also foster ethical, critical, and adaptive thinking among students. This study explores the implementation of a staggered and collaborative learning approach for teaching AI-assisted academic writing in the Software Measurement and Quality course at Universitas Ma Chung. The instructional design began with the formation of small student groups, followed by a structured introduction to generative AI tools—such as ChatGPT Plus and Consensus—paired with discussions on ethical use, fact-checking, and reference validation. Rather than emphasizing technical details, the concept of prompt engineering was taught as a set of adaptable, strategic principles, enabling students to construct effective prompts as technology evolves. Through staggered, step-by-step activities, students were guided from initial brainstorming and article drafting to in-depth literature reviews and iterative revision, with each group collaborating internally and across teams to share references and insights using tools like Mendeley Reference Manager. The lecturer’s ongoing formative feedback, combined with originality checks via Turnitin, ensured both academic integrity and continuous improvement. At the end of the process, students submitted their articles to designated journals and completed a feedback questionnaire. The analysis of student responses revealed overwhelmingly positive outcomes: nearly all participants reported a clearer understanding of course objectives, increased motivation, and greater creativity. Students highlighted the ease and effectiveness of AI tools in supporting their research, the value of collaborative reference sharing, and the importance of ethical AI practices. The majority also recognized that prompt engineering requires strategic, iterative approaches rather than fixed formulas. In conclusion, the findings demonstrate that the staggered and collaborative learning model not only enhances academic writing skills and ethical awareness but also prepares students for lifelong learning and responsible technology use in an AI-driven academic landscape.