Article Per Year (5 Year)
p-Index From 2021 - 2026
1.006
P-Index
This Author published in this journals
All Journal Sistemasi: Jurnal Sistem Informasi Indonesian Journal of Artificial Intelligence and Data Mining JIPI (Jurnal Ilmiah Penelitian dan Pembelajaran Informatika) Jurnal Triton JuTISI (Jurnal Teknik Informatika dan Sistem Informasi) Journal of Information Technology JISTECH: Journal of Information Science and Technology
Alex De Kweldju
Universitas Papua
Agriculture, Biological Sciences & Forestry Biochemistry, Genetics & Molecular Biology Computer Science & IT Education Electrical & Electronics Engineering Environmental Science Social Sciences Veterinary

Published : 7 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 7 Documents
Search

 1 
Sistem Manajemen Pembelajaran Lokal untuk Meningkatkan Pemahaman Belajar Mahasiswa Fridolin Febrianto Paiki; Alex De Kweldju; Ratna Juita
Jurnal Teknik Informatika dan Sistem Informasi Vol 6 No 1 (2020): JuTISI
Publisher : Maranatha University Press

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.28932/jutisi.v6i1.1818

Abstract

Sistem Manajemen Pembelajaran merupakan salah satu perangkat penting yang digunakan di sekolah dan kampus untuk meningkatkan kemampuan dan hasil studi siswa. Ketersediaan server web dan jaringan komputer yang cepat, handal, murah, dan mudah diakses serta dipadukan dengan Sistem Manajamen Pembelajaran atau Learning Management System (LMS) dapat membantu mahasiswa untuk mengikuti proses perkuliahan dengan lebih baik. Web server diimplementasikan menggunakan virtual machine (VM) Proxmox pada komputer di laboratorium komputer di program studi. Komputer yang digunakan memiliki spesifikasi standar (bukan server), seperti Intel Processor i3, HDD 256GB, dan memori 2GB. Kualitas web server diukur menggunakan Apache Bench dan Web Server Stress Tool. Hasil simulasi menggunakan Apache Bench diperoleh waktu yang diperlukan untuk menyelesaikan 1.000 concurrent request adalah sebanyak 13.823 milidetik. Meskipun belum optimal, hasil ini sudah mendekati kondisi ideal, yaitu kurang dari 10 detik. Server diimplementasikan secara virtual agar dapat dengan mudah dideploy ke dalam server lokal lainnya di fakultas-fakultas lain. Hasil penelitian menunjukkan bahwa kinerja server masih perlu ditingkatkan, baik dengan meningkatkan kinerja agar tercapai kondisi ideal mengingat kebutuhan mahasiswa dan dosen akan terus meningkat. Kualitas jaringan pun perlu diperhatikan, seperti penggunaan firewall, antivirus, dan aplikasi lainnya yang bisa mempengaruhi kualitas respon yang diterima oleh user. Diversifikasi media pembelajaran juga perlu digunakan agar hasil penelitian dapat juga mencakup kondisi dan protokol yang berbeda.
Aplikasi Media Pembelajaran Tata Surya Untuk Anak Sekolah Dasar Menggunakan Augmented Reality Berbasis Android : Aplikasi Media Pembelajaran Tata Surya Untuk Anak Sekolah Dasar Menggunakan Augmented Reality Berbasis Android Sutimin Sutimin; Julius Panda Putra Naibaho; Alex De Kweldju
JISTECH: Journal of Information Science and Technology Vol 13 No 2 (2024): Volume 13 Nomor 2 Tahun 2024
Publisher : Universitas Papua

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30862/jistech.v13i2.443

Abstract

Along with the development of information technology which continues to move from year after year, many learning methods have been created to support teaching and learning process. Learning media is in the form of learning props. The solar system currently still uses print media, videos and teaching aids another simple one where the teacher is more dominant in explaining and the students are the only ones listening. Methods like this don't hone creativity and grasping power students, plus the limited use of teaching aids, on the other hand their availability the latest technology that can be developed into learning media. Method learning with the support of interesting technology-based media will creating effective learning media to help achieve targets learning. Application of augmented reality in learning materials about The solar system provides innovative learning methods that can create interactive learning communication between teachers and students. Method The research method used is the Research and method Development (R&D) which aims to create applied products. In the process of creating this application, the application development process used at this stage using the waterfall developer model. The method is carried out is to do Augmented Reality with Vuforia and Unity to implement on Android devices. The result achieved was the introduction of the solar system displays objects and information. With the Augmented Solar System application Android-based reality can be useful for visualizing the inner solar system more realistic model shape.
Analisis Integrasi Algoritma YOLOv8 dan CNN terhadap Klasifikasi Kualitas Telur Ayam Ras Petelur di Kabupaten Manokwari Helmi Saputra; Christian Dwi Suhendra; Alex De Kweldju
JURNAL TRITON Vol 17 No 1 (2026): JURNAL TRITON
Publisher : Politeknik Pembangunan Pertanian Manokwari

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47687/jt.v17i1.1912

Abstract

Penilaian mutu telur ayam petelur umumnya masih dilakukan secara manual sehingga rentan terhadap subjektivitas dan ketidakkonsistenan hasil, sementara kebutuhan akan proses grading yang cepat dan akurat semakin meningkat. Penelitian ini dilakukan untuk mengembangkan model klasifikasi mutu telur berbasis warna dan bentuk kerabang dengan memanfaatkan algoritma YOLOv8 dan Convolutional Neural Network (CNN). Data dikumpulkan dari peternak lokal di Kabupaten Manokwari menggunakan metode Stratified Random Sampling dan menghasilkan 957 citra telur yang diberi label berdasarkan pedoman SNI 3926:2023. Proses preprocessing meliputi penyesuaian ukuran citra menjadi 640×640 piksel, normalisasi, serta augmentasi variasi pencahayaan dan sudut. YOLOv8 digunakan untuk mendeteksi posisi telur, sementara EfficientNet-B0 sebagai CNN digunakan untuk klasifikasi tiga tingkat mutu telur. Hasil penelitian menunjukkan bahwa YOLOv8 mampu mendeteksi objek dengan performa sangat baik dengan nilai mAP@0.5 sebesar 0.980, precision 0.97, dan recall 0.95, sedangkan model CNN mencapai akurasi validasi 91.9% dengan performa stabil pada seluruh kelas. Integrasi kedua model menghasilkan sistem yang mampu bekerja secara real-time dengan kecepatan 20–35 FPS dan dapat menghitung jumlah telur per grade secara otomatis. Berdasarkan hasil tersebut, dapat disimpulkan bahwa kombinasi YOLOv8 dan CNN mampu memberikan solusi yang cukup akurat untuk proses grading telur ayam petelur serta berpotensi diterapkan pada sistem sortasi pascapanen skala produksi.
Analysis of Factors Influencing the Adoption of the PELNI Mobile Application for Ship Ticket Purchases in Sorong using the Unified Theory of Acceptance and Use of Technology 2 Regina Natalia Sianturi; Julius Panda Putra Naibaho; Alex De Kweldju
SISTEMASI Vol 15, No 5 (2026): Sistemasi: Jurnal Sistem Informasi
Publisher : Universitas Islam Indragiri

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.32520/stmsi.v15i5.5866

Abstract

This study aims to analyze the factors influencing the adoption of the PELNI Mobile application for ship ticket purchases using the Unified Theory of Acceptance and Use of Technology 2 (UTAUT2) approach. The research employed a quantitative method with data collected through questionnaires distributed to users of the PELNI mobile application. The collected data were analyzed using Structural Equation Modeling–Partial Least Squares (SEM-PLS). The results indicate that Facilitating Conditions and Habit significantly influence the actual usage behavior of the PELNI mobile application. In contrast, Performance Expectancy, Effort Expectancy, Social Influence, Hedonic Motivation, and Price Value do not have a significant effect on usage intention. Furthermore, Behavioral Intention was also found to have no significant influence on actual usage behavior. These findings suggest that the adoption of the PELNI mobile application in the context of maritime transportation is more strongly influenced by the availability of supporting facilities and users’ habitual behavior than by perceived usefulness, ease of use, or social influence. This study is expected to contribute theoretically to the development of the UTAUT2 model and provide practical implications for the management of digital-based maritime transportation services.
Classification of Wild Edible Plants Using InceptionV3 with Transfer Learning and Metadata Integration as a Decision Support System Ridho Nur Fauzi; Julius Panda Putra Naibaho; Alex De Kweldju
Indonesian Journal of Artificial Intelligence and Data Mining Vol. 9 No. 1 (2026): March 2026
Publisher : Universitas Islam Negeri Sultan Syarif Kasim Riau

Show Abstract | Download Original | Original Source | Check in Google Scholar

Abstract

Deep learning has advanced intelligent systems for plant identification; however, distinguishing edible wild plants remains challenging due to limited datasets and the need for contextual information beyond visual classification. This study develops a Convolutional Neural Network (CNN) framework that integrates metadata as a decision support system to enhance food safety and strengthen community-based food security. A dataset of 16,076 images across 34 classes of edible wild plants was collected and enriched with metadata containing plant descriptions, consumption status, and nutritional values. The dataset was split into 75% training, 20% validation, and 5% testing to ensure reliable evaluation. The proposed solution employs InceptionV3 with transfer learning as the primary model, chosen for its ability to capture complex visual features in limited datasets, while MobileNetV3-Large serves as a lightweight comparative architecture. Results show that InceptionV3 achieved superior performance with a test accuracy of 0.87 and F1-score of 0.88, whereas MobileNetV3-Large obtained only 0.03 accuracy, indicating poor generalization. This highlights the importance of selecting architectures with sufficient depth for domains characterized by high visual variability. Metadata integration enhanced the system’s role as a decision support tool, providing contextual information such as edibility status and nutritional content. The novelty of this research lies in combining CNN-based classification with metadata integration, transforming the system into a practical framework for safe consumption decisions. Limitations include the dataset containing only edible plants. Future work should incorporate non-edible classes, evaluate performance under real-world conditions, and explore advanced architectures and explainable AI techniques to improve robustness, transparency, and accessibility.
Clustering HIV Screening Data in Teluk Bintuni Using K-Means Yuliana Manobi; Alex De Kweldju; Julius Panda Putra Naibaho
J-INTECH ( Journal of Information and Technology) Vol 14 No 02 (2026): Journal of Information and Technology
Publisher : LPPM Universitas Bhinneka Nusantara

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.32664/j-intech.v14i02.2368

Abstract

Human Immunodeficiency Virus (HIV) remains a major public health challenge in Papua, Indonesia, where geographical barriers and limited resources complicate service delivery. This study applies clustering methods to HIV screening data from 22 Health Service Units (UPK) in Teluk Bintuni Regency during 2024–2025. The final dataset consisted of 29 valid unit-year observations containing variables related to total tests, HIV-positive cases, and sex-disaggregated distributions. Data preprocessing followed the Knowledge Discovery in Database (KDD) framework, including data selection, cleaning, log transformation, normalization, and the construction of derived variables such as positivity rate and gender ratios. Four clustering algorithms were compared, namely K-Means, Hierarchical Clustering, Fuzzy C-Means, and DBSCAN, using Silhouette Score, Davies-Bouldin Index, and Calinski-Harabasz Index. The results indicate that K-Means produced the most stable and interpretable clustering structure, forming three groups of UPKs: intermediate screening units with moderate coverage and low positivity, priority units with limited testing but high positivity rates, and active screening units with the highest testing volume and case detection. These findings reveal heterogeneity in HIV screening performance across UPKs and support differentiated intervention strategies. Units with high apparent positivity but low testing coverage require expanded outreach, field verification, and improved access to HIV screening services, while active screening units should be strengthened through counseling, referral, and follow-up services. This study demonstrates the usefulness of clustering analysis in identifying service gaps and supporting evidence-based HIV intervention planning at the local level.
PENERAPAN ALGORITMA BERT UNTUK DETEKSI KOMENTAR SPAM JUDI ONLINE DI YOUTUBE Fahri Akbar Rosid Asro; Julius Panda Putra Naibaho; Alex De Kweldju
JIPI (Jurnal Ilmiah Penelitian dan Pembelajaran Informatika) Vol 11, No 2 (2026)
Publisher : STKIP PGRI Tulungagung

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.29100/jipi.v11i2.8122

Abstract

Penyebaran komentar spam yang mengandung promosi perjudian daring di YouTube semakin meningkat dan mengganggu kualitas interaksi digital. Deteksi manual tidak lagi efektif karena volume komentar yang tinggi dan kompleksitas gaya bahasa yang digunakan. Penelitian ini bertujuan untuk menerapkan model In-doBERT yang telah di‑fine‑tune guna mendeteksi komentar spam bertema perjudian dalam bahasa Indonesia. Data mentah diperoleh melalui scraping sebanyak 9.056 komentar dari 14 video kategori game. Setelah dilakukan pra‑pemrosesan yang meliputi pembersihan teks, normalisasi, tokenisasi, penghapusan stopword, dan stemming, 263 komentar dihapus sehingga tersisa 8.793 komentar bersih untuk pelabelan. Komentar dikategorikan secara semi‑otomatis ke dalam dua kelas, spam dan non‑spam, dengan bantuan pencocokan kata kunci dan validasi manual. Model IndoBERT dilatih menggunakan pendekatan transfer learning dan dievaluasi dengan metrik akurasi, presisi, recall, dan F1‑score. Hasil eksperimen menunjukkan akurasi sekitar 99,54% (atau dibulatkan menjadi 99,6%) dan F1‑score 1,00, dengan hanya delapan kesalahan klasifikasi dari 1.759 data uji. Vis-ualisasi melalui wordcloud dan histogram mengungkap pola domi-nan istilah promosi dalam komentar spam. Temuan ini menunjukkan bahwa pendekatan berbasis transformer efektif dalam memahami konteks semantik, meskipun disamarkan melalui eufemisme atau kode, serta dapat diandalkan untuk meningkatkan akurasi sistem moderasi otomatis pada platform digital.
Co-Authors Christian Dwi Suhendra Fahri Akbar Rosid Asro Fridolin Febrianto Paiki Helmi Saputra Julius Panda Putra Naibaho Julius Panda Putra Naibaho Ratna Juita Regina Natalia Sianturi Ridho Nur Fauzi Sutimin Sutimin Yuliana Manobi