Article Per Year (5 Year)
p-Index From 2021 - 2026
3.708
P-Index
This Author published in this journals
All Journal International Journal of Advances in Applied Sciences Jurnal Teknologi Informasi dan Ilmu Komputer JTERA (Jurnal Teknologi Rekayasa) Jurnal SOLMA Pendas : Jurnah Ilmiah Pendidikan Dasar ILKOM Jurnal Ilmiah Jurnal Nasional Pendidikan Teknik Informatika (JANAPATI) Mobile and Forensics Zonasi: Jurnal Sistem Informasi Indonesian Journal of Data and Science Jurnal Algoritma, Logika dan Komputasi Jurnal Teknologi Sistem Informasi Buletin Sistem Informasi dan Teknologi Islam Jurnal Pustaka AI : Pusat Akses Kajian Teknologi Artificial Intelligence Jurnal Informatika Progres Internet of Things and Artificial Intelligence Journal Journal of Computer Science and Information Technology ARUS JURNAL SAINS DAN TEKNOLOGI
Desi Anggreani
Universitas Muhammadiyah Makassar
Humanities Civil Engineering, Building, Construction & Architecture Computer Science & IT Control & Systems Engineering Decision Sciences, Operations Research & Management Earth & Planetary Sciences Economics, Econometrics & Finance Education Electrical & Electronics Engineering Engineering Environmental Science Industrial & Manufacturing Engineering Languange, Linguistic, Communication & Media Library & Information Science Materials Science & Nanotechnology Mathematics Mechanical Engineering Neuroscience Physics Social Sciences Other

Published : 32 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 32 Documents
Search

 1   2   3   4 
IMPLEMENTASI MODEL TRANSFORMER INDOBERT BERBASIS DEEP LEARNING UNTUK DETEKSI SPAM PESAN TEKS BERBAHASA INDONESIA Annisa Alfrini; Desi Anggreani; Fahrim Irhamna Rachman
Journal of Computer Science and Information Technology Vol. 3 No. 3 (2026): Juni
Publisher : Yayasan Nuraini Ibrahim Mandiri

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.70248/jcsit.v3i3.4147

Abstract

Perkembangan teknologi informasi dan komunikasi dalam beberapa tahun terakhir telah mendorong peningkatan penggunaan media digital secara masif, salah satunya adalah aplikasi Telegram. Namun, fenomena ini memunculkan permasalahan baru, yaitu ancaman penyebaran pesan spam yang tidak terkendali, berpotensi mengganggu kenyamanan, menyebabkan kerugian finansial, serta mengancam keamanan data pengguna akibat skema penipuan. Penelitian ini bertujuan untuk mengimplementasikan model deep learning berbasis arsitektur Transformer, secara spesifik menggunakan model IndoBERT, untuk mendeteksi dan mengklasifikasikan pesan teks Telegram berbahasa Indonesia ke dalam kategori spam dan non-spam (ham). Penelitian ini memanfaatkan dataset seimbang yang berjumlah 5.971 pesan teks (2.979 spam dan 2.992 non-spam), yang dikumpulkan secara langsung melalui metode scraping dari berbagai grup Telegram publik. Tahapan penelitian dirancang secara sistematis, meliputi prapemrosesan teks (case folding, cleaning, dan tokenisasi awal), tokenisasi lanjutan menggunakan metode WordPiece, penyesuaian parameter, serta proses fine-tuning pada model pralatih IndoBERT. Evaluasi performa model diukur secara komprehensif menggunakan confusion matrix. Hasil penelitian menunjukkan bahwa model IndoBERT berhasil mengidentifikasi pola linguistik spam, termasuk penggunaan bahasa informal, tautan eksternal, hingga manipulasi karakter. Model menghasilkan kinerja klasifikasi yang sangat optimal dengan perolehan nilai akurasi sebesar 97,32%, presisi 98,62%, recall 95,97%, dan F1-score 97,28%. Selain itu, pengujian langsung terhadap 10 sampel data baru (unseen data) menunjukkan tingkat keberhasilan prediksi sebesar 100% secara spesifik pada skala pengujian tersebut. Kesimpulannya, pendekatan model IndoBERT terbukti sangat efektif, tangguh, dan adaptif untuk diimplementasikan sebagai sistem penyaringan spam otomatis secara real-time pada teks berbahasa Indonesia.
A Hybrid Salp Swarm Optimization and Behavioral Nudge Framework for Optimizing Software Developer Task Allocation Ashabul Kahfi; Muhammad Faisal; Titin Wahyuni; Desi Anggreani; Darniati Darniati; Muhammad Syafaat S Kuba; Andi Makbul Syamsuri; Ida Mulyadi
Jurnal Nasional Pendidikan Teknik Informatika: JANAPATI Vol. 15 No. 2 (2026)
Publisher : Prodi Pendidikan Teknik Informatika Universitas Pendidikan Ganesha

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.23887/janapati.v15i2.106920

Abstract

Effective task allocation is critical in Agile software development, yet most optimization-based approaches treat it as a purely technical scheduling problem and disregard behavioral factors such as motivation, fairness, and engagement. This study proposes a Hybrid Salp Swarm Optimization–Behavioral Nudge Framework (HSSO–BNF) for developer–task allocation that integrates technical constraints with human-centered cues. The model formulates allocation as a multi-objective function combining workload balance, skill mismatch, deadline penalties, and a motivation score derived from three nudge components: Motivational Cue (MC), Social Comparison (SC), and Effort–Reward Feedback (ERF). These behavioral signals are embedded directly into the SSO position update and fitness evaluation, enabling the swarm to adapt simultaneously to performance and motivational states. Experiments on real developer–task records collected from GitHub compare HSSO–BNF against GA, PSO, and standard SSO using convergence behavior, allocation cost, fairness, satisfaction, and motivation dynamics. The results show that HSSO–BNF achieves faster and more stable convergence, reduces allocation cost by approximately 32% compared with GA and SSO and about 25% compared with PSO, and improves workload fairness and developer satisfaction while preserving psychologically sustainable specialization patterns. Heatmap visualizations and motivation trends further confirm that the behavioral layer produces more coherent and interpretable task assignments, indicating that behavior-aware metaheuristics are a promising direction for intelligent, human-centered task allocation in Agile teams.
PERBANDINGAN METODE ANFIS DAN ANN DALAM KLASIFIKASI PENYAKIT STROKE MENGGUNAKAN DATA REKAM MEDIS PASIEN Hidayah, Nur; Anggreani, Desi; Danuputri, Chyquitha
Pendas : Jurnal Ilmiah Pendidikan Dasar Vol. 11 No. 03 (2026): Volume 11 Nomor 03, September 2026 Completed
Publisher : Program Studi Pendidikan Guru Sekolah Dasar FKIP Universitas Pasundan

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.23969/jp.v11i03.63396

Abstract

Accurate stroke classification is vital for appropriate clinical treatment. This study compares the Adaptive Neuro-Fuzzy Inference System (ANFIS) and Artificial Neural Network (ANN) in classifying ischemic and hemorrhagic stroke using 368 medical records from RSKD Dadi, South Sulawesi. Data were preprocessed using Min-Max normalization and an 80:20 train-test split. ANFIS used Fuzzy C-Means clustering with five fuzzy rules, while ANN used a Multi-Layer Perceptron (10-14-7-2). Evaluated via Confusion Matrix, ANFIS outperformed ANN across all metrics, achieving 82.43% Accuracy, 82.00% Precision, 78.00% Recall, and 79.00% F1-Score, compared to ANN’s 79.73%, 80.00%, 73.00%, and 75.00%. ANFIS is a more effective and interpretable model for supporting stroke diagnosis.
Optimalisasi Automated Unit Testing pada Pemrograman Berbasis Web Menggunakan Pendekatan Reinforcement Learning Muh Akram Riyadi Ramadhan; Muhammad Faisal; Lukman Lukman; Desy Anggreani; M. Agusalim; Irnawaty Idrus; Soemitro Emin Praja
Buletin Sistem Informasi dan Teknologi Islam (BUSITI) Vol 7, No 3 (2026)
Publisher : Universitas Muslim Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33096/busiti.v7i3.3507

Abstract

Pengujian perangkat lunak penting dalam pengembangan aplikasi web, namun automated unit testing konvensional berbasis aturan statis terbatas menghadapi kompleksitas sistem modern. Penelitian ini memformulasikan prioritisasi kasus uji sebagai Markov Decision Process: agen Reinforcement Learning mengamati state, memilih kasus uji, dan menerima reward multiobjektif yang memadukan cakupan kode dan deteksi kesalahan dikurangi biaya eksekusi. Tiga algoritma (DQN, DDQN, PPO) dibandingkan terhadap baseline Random dan Greedy menggunakan lima seed dan uji Mann–Whitney U. Pada System Under Test berbasis Flask, DDQN memperoleh performa terbaik dengan APFD 0,8504, NAPFD 0,8447, dan F1 0,4183, mengungguli baseline secara signifikan (nilai p terkecil 7,15 × 10⁻¹⁴⁴) dengan latensi keputusan 3,0033 ms, kurang dari separuh PPO. Greedy mencapai cakupan tertinggi namun deteksi terendah, menegaskan cakupan bukan indikator memadai. Validasi eksternal pada BugsInPy (82 bug web) mengonfirmasi arah keunggulan serupa dengan margin lebih kecil. Reinforcement Learning merupakan pendekatan menjanjikan dan dapat direproduksi untuk automated unit testing adaptif pada aplikasi web.
Penerapan Aplikasi Keuangan Berbasis Web Terhadap Peningkatan Efektivitas dan Transparansi dalam Mengelola Keuangan Sekolah Ismawati; Desi Anggreani; Widia Wahyuni; Ikha Andriani
Jurnal SOLMA Vol. 15 No. 2 (2026)
Publisher : Universitas Muhammadiyah Prof. DR. Hamka (UHAMKA Press)

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

Abstract

Background: Pengelolaan keuangan di SD UNISMUH Makassar masih menghadapi kendala signifikan karena mengandalkan sistem pencatatan manual menggunakan buku kas dan spreadsheet sederhana. Sistem manual ini memicu berbagai permasalahan, antara lain waktu pemrosesan rekapitulasi laporan bulanan yang mencapai 5-7 hari kerja, tingkat kesalahan input sebesar 15%, serta data keuangan yang tersebar dan tidak terintegrasi. Selain itu, kurangnya kompetensi sumber daya manusia dan ketidakseragaman penggunaan standar nama akun maupun format laporan menyebabkan informasi keuangan menjadi kurang akurat dan tidak transparan. Akibatnya, pengambilan keputusan operasional terkait pengelolaan dana dan sumber daya manusia oleh pihak sekolah menjadi terhambat. Oleh karena itu, diperlukan inovasi berupa transformasi ke sistem pengelolaan keuangan otomatis berbasis web untuk meningkatkan efisiensi, akurasi, dan transparansi laporan keuangan sekolah. Tujuan: Mengimplementasikan dan mendampingi penerapan sistem pengelolaan keuangan otomatis berbasis web (aplikasi KeuanganDigitalMuh.id) di SD UNISMUH Makassar untuk menggantikan sistem pencatatan manual. Metode: Metode pelaksanaan yang digunakan disusun secara partisipatif dengan mengajak mitra berdiskusi untuk menentukan solusi yang paling sesuai yaitu melalui Perancangan Sistem Terautomasi, Pelatihan Sumber daya manusia, Standardisasi Nama Akun dan Format Laporan, Pendampingan dan Evaluasi Keberlanjutan. Hasil: Hasil dari kegiatan pengabdian ini mencakup dua pencapaian utama: keberhasilan perancangan dan implementasi sistem aplikasi keuangan terautomasi berbasis web yang disesuaikan dengan kebutuhan mitra; dan peningkatan kompetensi sumber daya manusia melalui pelatihan dan pendampingan langsung kepada guru dan staf. Kesimpulan: Implementasi dan pelatihan ini secara efektif berhasil mengalihkan mitra dari pencatatan manual ke sistem digital, yang berdampak pada efisiensi waktu, standarisasi format laporan, serta peningkatan akurasi dan transparansi keuangan sekolah.
Model Hibrida PCA-MLP dan Optimasi untuk Klasifikasi Status Gizi Balita Berbasis Fitur Antropometri: Studi Kasus Kelurahan Tamarunang, Sulawesi Selatan Tenri Batari, Andi Firqatun Najiah; Anggreani, Desi; Lukman
Jurnal Teknologi Informasi dan Ilmu Komputer Vol 13 No 4: Agustus 2026
Publisher : Fakultas Ilmu Komputer, Universitas Brawijaya

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.25126/jtiik.134

Abstract

Stunting merupakan masalah gizi kronis yang berdampak jangka panjang terhadap kualitas kesehatan dan perkembangan anak, sehingga diperlukan pendekatan deteksi dini yang akurat dan berbasis data. Penelitian ini bertujuan untuk mengembangkan dan mengevaluasi model machine learning dalam mengklasifikasikan status gizi balita berdasarkan data antropometri. Dataset yang digunakan terdiri atas 638 balita dari Kelurahan Tamarunang, Sulawesi Selatan, dengan enam variabel prediktor antropometri yaitu berat badan lahir, tinggi badan lahir, usia saat ukur berat, tinggi dan LiLA. Metodologi penelitian meliputi seleksi fitur menggunakan Mutual Information (MI), penyeimbangan data menggunakan Synthetic Minority Over-sampling Technique (SMOTE), serta pemodelan menggunakan Multilayer Perceptron (MLP), Random Forest, dan model hibrida Principal Component Analysis–Multilayer Perceptron (PCA-MLP). Hasil analisis MI menunjukkan bahwa berat badan (MI = 0,053) dan tinggi badan (MI = 0,038) saat pengukuran merupakan fitur paling informatif. Evaluasi komparatif menunjukkan bahwa model hibrida PCA-MLP mencapai kinerja terbaik dengan akurasi 93,70% dan F1-Score (Macro) sebesar 0,551. Meskipun demikian, analisis per kelas mengungkapkan bahwa deteksi kelas minoritas masih menjadi tantangan akibat ketidakseimbangan data. Oleh karena itu, model yang diusulkan berpotensi digunakan sebagai alat pendukung keputusan untuk skrining awal status gizi, namun tidak dimaksudkan sebagai pengganti penilaian klinis oleh tenaga kesehatan.   Abstract Stunting is a chronic nutritional problem that hinders children's development and requires an accurate early detection system. This study aims to develop and evaluate a high-performing machine learning model for classifying the nutritional status of under-five children. The research utilized data from 638 children in Tamarunang Village, Gowa Regency, incorporating six anthropometric predictor variables. The proposed methodology includes feature selection using Mutual Information (MI), data balancing with Synthetic Minority Over-sampling Technique (SMOTE), and modeling using three architectures: Multilayer Perceptron (MLP), Random Forest, and a hybrid Principal Component Analysis-Multilayer Perceptron (PCA-MLP) model. MI analysis results indicate that Body Weight (MI=0.053) and Height (MI=0.038) at measurement exhibit the highest statistical relevance. In comparative evaluation, the Hybrid PCA-MLP model demonstrated the most superior performance with an accuracy of 93.70% and an F1-Score (Macro) of 0.551, outperforming Random Forest (0.416) and MLP (0.320). Visualization via t-SNE and per-class metrics confirmed that minority class classification remains a primary challenge due to data imbalance. This research affirms that a hybrid approach combined with appropriate feature selection holds significant potential as a decision support tool for malnutrition screening.
Swin Transformer Enhanced with OOD Detection for Robust and Reliable Diagnosis of Ischemic Stroke from CT Image Nurmisba Nurmisba; Desi Anggreani; Muhyiddin A M Hayat; Aedah Abd Rahman
Jurnal Nasional Pendidikan Teknik Informatika: JANAPATI Vol. 14 No. 3 (2025)
Publisher : Prodi Pendidikan Teknik Informatika Universitas Pendidikan Ganesha

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.23887/janapati.v14i3.102864

Abstract

Diagnosing ischemic stroke from CT scan images presents significant challenges in achieving the speed and accuracy essential for clinical decision-making, where conventional CNN-based methods show limitations. This study addresses these gaps by developing an automated diagnostic system using a Swin Transformer model integrated with an Out-of-Distribution (OOD) detection mechanism to enhance diagnostic reliability. The model was trained and validated on a dataset of 583 brain CT images from 341 patients at a regional hospital in Makassar. This dataset, labeled by two expert radiologists (κ=0.94), was categorized into ischemic stroke (206), normal (228), and non-brain CT scans (149) as the OOD class. The Swin Transformer achieved an exceptional validation accuracy of 99.15% after 10 epochs, with a highly efficient total training time of approximately 24 minutes. The model’s superiority was further confirmed by high weighted averages for precision (0.99), recall (0.99), and F1-score (0.99). Critically, the OOD detection module demonstrated perfect performance, achieving 100% accuracy in identifying irrelevant images with a 0% false positive rate, thereby preventing erroneous diagnoses from non-brain scans. Robustness testing under varied lighting conditions also showed a 100% success rate. Real-time viability was confirmed through external validation using a live camera, yielding a rapid inference time of 0.3 seconds per image. This study concludes that the developed system offers a highly accurate, robust, and safe solution, proving its readiness for clinical implementation to support ischemic stroke diagnosis in Indonesia.
T4EDR: Hybrid Threat Detection Framework for EDR Based on Semantic Rule Embedding and Contextual Network Flow Analysis Muh Dzikri Alfauzan Nuzul; Desi Anggreani; Muhammad Faisal; Abd Rahman Wahid; Titik Khawa Abd Rahman
Jurnal Nasional Pendidikan Teknik Informatika: JANAPATI Vol. 15 No. 1 (2026)
Publisher : Prodi Pendidikan Teknik Informatika Universitas Pendidikan Ganesha

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.23887/janapati.v15i1.104059

Abstract

In response to the increasing complexity of cyber threats and attacks, this study proposes an innovative endpoint detection and response (EDR) framework named T4EDR (Transformer for Endpoint Detection and Response). Specifically, it addresses the inefficiencies of traditional systems that rely on static and less adaptive rules. T4EDR integrates semantic analysis methods to interpret security rules and a transformer-based deep learning model to contextually analyze network traffic flows. In the initial phase, thousands of security rules from Wazuh were extracted, embedded, and semantically validated against the MITRE ATT&CK framework, achieving a semantic coherence of 86.43% with a silhouette score of 0.702. Subsequently, the FlowBERT model was designed to classify network traffic flows using the CIC-IDS2018 dataset, achieving 91.1% accuracy, a macro-F1 of 0.79, and a mean Average Precision (mAP) of 0.90, surpassing the quantitative target of 85%. The integration of rule embeddings with FlowBERT hidden states through a linear projector enables adaptive mapping of endpoint activities to relevant security rules, supporting context-based automated responses. The main contribution of this study is an adaptive framework that bridges the gap between traditional rule analysis and deep learning-based detection, thereby enhancing the capability to detect multi-stage threats on modern endpoints.
OPTIMASI KLASIFIKASI JENIS INDUSTRI KECIL MENENGAH (IKM) MENGGUNAKAN DEEP NEURAL NETWORK BERBASI SHAP ANALYSIS M. Fikri Haikal Ayatullah; Desi Anggreani; Rizki Yusliana Bakti; Muhammad Faisal; Muhammad Syafaat; M. Agusalim
PROGRESS Vol 18 No 2 (2026): September
Publisher : P3M STMIK Profesional Makassar

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.56708/progres.v18i2.463

Abstract

Small and Medium Industries (SMEs) in Makassar City face a significant gap between high labor absorption (66.25%) and low GDP contribution (20%), often due to conventional and experience-based determination of industry types. This study implements a Deep Neural Network (DNN) model to classify four categories of SMEs, Bread and Cake, Processed Food, Vehicle Repair, and Textiles and Clothing to facilitate data-driven decisions. Using a supervised learning approach on 31,824 data samples for the 2022-2024 period, this model was developed through feedforward and backpropagation mechanisms. The results showed superior performance with overall accuracy of (93.99%) and balanced accuracy (96.70%), which signified an increase of (15.88%) compared to the Naïve Bayes baseline model. All F1-scores above (90%) indicate strong performance stability in each class. Furthermore, SHAP's analysis revealed that textual features (87.1%) were the dominant factors, followed by the type of business entity and investment value. This study confirms that DNN is effective in modeling complex non-linear interactions, providing objective tools for classification and strategic economic planning in Makassar City.
KLASIFIKASI MULTI-CLASS STATUS GIZI BALITA MENGGUNAKAN ARSITEKTUR DEEP NEURAL NETWORK Alizha Nur Arspandy; Desi Anggreani; Muhyiddin A.M Hayat; Muhammad Faisal; Muhammad Syafaat; Indriyanti; Emil Aguslaim Habi Thalib
PROGRESS Vol 18 No 2 (2026): September
Publisher : P3M STMIK Profesional Makassar

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.56708/progres.v18i2.464

Abstract

This study aims to develop a classification model for toddler nutritional status using a Deep Neural Network (DNN) with a multi-class classification approach. The research utilizes anthropometric data of toddlers aged 0-60 months obtained from UPTD Puskesmas Cendana Putih, North Luwu Regency, covering the period 2023–2025. The dataset consists of 156 records with features including age, weight, height, and Z-score indicators. Data preprocessing involves validation, normalization, and splitting into training and testing sets with a ratio of 85:15. The DNN model is constructed with multiple hidden layers (128, 64, and 32 neurons) and trained using the Adam optimizer and categorical cross-entropy loss function. The results show that the model achieves an accuracy of 91.67% on the testing data, indicating good performance in classifying nutritional status into categories such as undernutrition, normal, and obesity. Evaluation using confusion matrix and classification metrics (precision, recall, and F1-score) reveals that the model performs well on dominant classes but shows limitations in minority classes due to data imbalance. Overall, the proposed model demonstrates potential as a decision support tool to assist healthcare workers in identifying toddler nutritional status more accurately and efficiently.
Co-Authors . Darniati A M Hayat, Muhyddin Abd Rahman Wahid Abd Rahman, Aedah Abd. Rasyid Syamsuri Abdul Rakhim Nanda Aedah Abd Rahman Aedah Abd Rahman Ahmad Faisal Aji Prasetya Wibawa Akbar, Muh Ilham Alizha Nur Arspandy Andi Citra Ayu Lestari Andi Inayah Auliyah Andi Makbul Syamsuri Annisa Alfrini Ashabul Kahfi Bakti, Rizki Yusliana Chyquitha Danuputri Chyquitha Danuputri Danuputri, Chyquitha Danuputri, Chyquitha Danuputri, Chyquitha Darniati Dedi Setiawan Dedi Setiawan Dewi MJ, Wanda Tyrana Emil Agus Salim Habi Talib Fachrim Irhamna Rachman Faturohman, Agung Galbi Nadifah Hamdani Herman Herman Ida Mulyadi, Ida Ikha Andriani Indriyanti Irnawaty Idrus Ismawati Ismawati Ismawati Khadijah Kusumawardani, Nurul Lukman Lukman Lukman Anas Lukman Lukman Lukman Lukman Lukman, Lukman LUKMAN, LUKMAN M Agusalim M. Fikri Haikal Ayatullah Muh Akram Riyadi Ramadhan Muh Dzikri Alfauzan Nuzul Muh Ilham Akbar Muhammad Faisal Muhammad Syafaat S. Kuba Muhyiddin A M Hayat Muhyiddin A.M Hayat Muhyiddin A.M Hayat Nur Hidayah Nur Milani Hidayah Nurmisba Nurmisba Nurmisba Nurmisba Nurmisba Nurmisba, Nurmisba Nurnawaty Nurul Kusumawardani Praja, Soemitro Emin Purnawansyah Purnawansyah Putri, Desy Pratiwi Ika Rahman, Aedah Abd Tenri Batari, Andi Firqatun Najiah Titik Khawa Abd Rahman Titin Wahyuni Try Gustaf Said Wahid, Abd Rahman Widia Wahyuni Widiyanti, Widiyanti Wiwik Handayani