Article Per Year (5 Year)
p-Index From 2021 - 2026
1.475
P-Index
This Author published in this journals
All Journal SMARTICS Journal The IJICS (International Journal of Informatics and Computer Science) JURIKOM (Jurnal Riset Komputer) Jurnal Algoritma Buffer Informatika
Fathoni
Universitas Sriwijaya
Computer Science & IT Control & Systems Engineering Decision Sciences, Operations Research & Management Electrical & Electronics Engineering

Published : 10 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 10 Documents
Search

 1 
Analisis UX E-PPT Universitas XYZ dengan Metode User Experience Questionnaire Kevin Agustin Purba Purba; Umar Rahman Zidan; Muhammad Azyumardi Azra; Fathoni
SMARTICS Journal Vol 11 No 2 (2025): Journal SMARTICS (Oktober 2025)
Publisher : Universitas PGRI Kanjuruhan Malang

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.21067/smartics.v11i2.11916

Abstract

Digital transformation in academic administrative services demands systems that are efficient and user-experience oriented. This study aims to analyze the user experience of the e-PPT website of the Faculty of Computer Science at XYZ University, which functions as an academic administrative service platform. The research employs the User Experience Questionnaire (UEQ) method, which measures six core aspects of user experience: attractiveness, efficiency, perspicuity, dependability, stimulation, and novelty. A total of 40 active students from the Faculty of Computer Science at XYZ University participated by completing the UEQ questionnaire. The analysis results indicate that all UEQ dimensions received low scores, with the novelty dimension scoring negatively. Benchmarking against similar systems revealed that the e-PPT website falls within the lowest quartile (bottom 25%) across all scales. These findings underscore the urgent need for comprehensive improvements in the website’s design and functionality to enhance the quality of digital academic services in the future.
Penilaian Risiko Fraud Transaksi Digital menggunakan Hybrid Machine Learning dengan Clustering dan Klasifikasi Hendra Wijaya; Naek Parulian Hutagalung; Mira Afrina; Ali Ibrahim; Fathoni
Jurnal Algoritma Vol 23 No 1 (2026): Jurnal Algoritma
Publisher : Institut Teknologi Garut

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33364/algoritma/v.23-1.3398

Abstract

Credit card transaction fraud detection is commonly treated as a binary classification problem, whereas operational risk management requires more detailed risk-level information to support investigation prioritization. This study proposes a hybrid machine learning framework for transaction risk stratification. In the first stage, the K-Means algorithm was applied to the training set to discover latent risk structures and generate cluster-based risk labels. Subsequently, a Random Forest model was trained to predict risk levels for new transaction data. To maintain evaluation objectivity, the dataset was divided into training, validation, and testing sets, and data leakage prevention mechanisms were implemented. The testing results show that the model was able to consistently classify two levels of risk with stable precision, recall, and F1-score values. In the binary fraud detection scenario, the model achieved an accuracy of 0.8831. These findings indicate that separating latent risk exploration from predictive classification can produce a more informative risk representation compared to conventional binary approaches. However, this study is still limited to a single public dataset and one classification model. Therefore, the generalizability and potential performance improvements of the model still need to be evaluated by experimenting with other algorithms.
Deteksi Dini Churn Pelanggan E-Commerce Berbasis CRM di Indonesia Menggunakan Algoritma Random Forest Alifa Putri Shahabiyah; Zakirah Sabrina Putri Pasha; Aliya Faiza; Ali Ibrahim; Fathoni
Buffer Informatika Vol. 12 No. 1 (2026): Buffer Informatika
Publisher : Department of Informatics Engineering, Faculty of Computer Science, University of Kuningan, Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.25134/buffer.v12i1.566

Abstract

Customer churn merupakan salah satu permasalahan utama dalam industri e-commerce karena berdampak langsung terhadap penurunan jumlah pelanggan dan pendapatan perusahaan. Meskipun berbagai penelitian telah dilakukan untuk meningkatkan akurasi prediksi churn, sebagian besar masih berfokus pada aspek teknis tanpa mengaitkannya secara langsung dengan implementasi dalam Customer Relationship Management (CRM). Oleh karena itu, penelitian ini bertujuan untuk membangun model deteksi dini churn pelanggan berbasis machine learning serta mengkaji pemanfaatannya dalam mendukung strategi CRM.Penelitian ini menggunakan dataset ulasan pelanggan e-commerce yang diolah melalui tahap preprocessing, transformasi data, dan pembentukan label churn. Model dibangun menggunakan algoritma Random Forest dengan pendekatan Cross-Validation. Hasil pengujian menunjukkan model berjalan sangat baik dengan akurasi mencapai 99,03%, presisi 76,92%, dan recall 84,42%.Hasil penelitian menunjukkan bahwa data ulasan pelanggan efektif dalam merepresentasikan perilaku pelanggan dan dapat digunakan untuk mendeteksi potensi churn sejak tahap awal interaksi. Model yang dihasilkan dapat diintegrasikan ke dalam sistem CRM untuk mendukung strategi retensi pelanggan secara proaktif melalui deteksi dini pelanggan berisiko churn.
Classification of Depression Severity Using a Random Forest Algorithm Based on Lifestyle, Demographic, and Psychological Factors haniyah faizah; Oktavio Theonady; Syalwa Salsabillah S; Fathoni; Ali Ibrahim
Buffer Informatika Vol. 12 No. 1 (2026): Buffer Informatika
Publisher : Department of Informatics Engineering, Faculty of Computer Science, University of Kuningan, Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.25134/buffer.v12i1.569

Abstract

Depression among college students is a mental health issue that impacts quality of life and academic performance. However, factors influencing depression levels such as lifestyle, demographics, and psychological factors have not yet been analyzed in an integrated manner. This study aims to develop a depression severity classification model using the Random Forest algorithm based on these factors. The dataset consists of 1,998 records with 16 features selected through the Knowledge Discovery in Database (KDD) process. To address data imbalance, the Synthetic Minority Over-sampling Technique (SMOTE) was applied. The results show that the Random Forest model achieved an accuracy of 97.88% and an AUC of 0.998. Feature importance analysis indicates that the variables Symptoms, Nervous Level, and Employment Status are dominant factors in determining depression levels. Based on these results, the model is capable of effectively classifying depression levels and has the potential to serve as the basis for an early detection system in the university setting.
Analisis Pola Keputusan Pembelian di Tokopedia berbasis Machine Learning untuk Customer Relationship Management Andin Sabilla Janna; Tamara Juliyanti; Muhammad Bayu Samudra; Ali Ibrahim; Fathoni
Buffer Informatika Vol. 12 No. 1 (2026): Buffer Informatika
Publisher : Department of Informatics Engineering, Faculty of Computer Science, University of Kuningan, Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.25134/buffer.v12i1.571

Abstract

Pertumbuhan pesat platform e-commerce telah memengaruhi keputusan pembelian konsumen, khususnya pada marketplace seperti Tokopedia. Penelitian ini bertujuan menganalisis pola keputusan pembelian produk makanan dan minuman serta merumuskan implikasinya terhadap strategi Customer Relationship Management (CRM) berbasis data. Metode yang digunakan adalah pendekatan kuantitatif dengan dataset sebanyak 9.711 ulasan pelanggan yang diperoleh melalui Kaggle. Analisis dilakukan menggunakan Naive Bayes untuk klasifikasi keputusan pembelian ulang dan FP-Growth untuk menemukan pola asosiasi antara rating, harga, dan keputusan pembelian. Tahapan praproses meliputi pembersihan data, transformasi atribut, diskretisasi, serta pemrosesan teks menggunakan TF-IDF. Hasil penelitian menunjukkan bahwa Naive Bayes menghasilkan akurasi 60,18% dengan precision kelas Y sebesar 94,30%, namun recall masih dipengaruhi oleh ketidakseimbangan data. FP-Growth menunjukkan bahwa rating bintang 5 merupakan faktor paling dominan dalam pembelian ulang, dengan kombinasi rating bintang 5 dan harga sedang menghasilkan confidence tertinggi sebesar 98,6%. Temuan ini menegaskan pentingnya kepuasan pelanggan dan mendukung perumusan strategi CRM yang lebih terarah.
The Impact of Dominant Color in the Shopee Application's User Interface Design on User Focus Levels Fathimah Fadiyah Salimah; Septi Mulia Putri; Badia Inaya Sazrade; Fathoni
The IJICS (International Journal of Informatics and Computer Science) Vol. 10 No. 2 (2026): July
Publisher : Universitas Budi Darma

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30865/ijics.v10i2.9686

Abstract

The design of mobile application interfaces plays a crucial role in shaping user experience, where color influences not only aesthetics but also user cognition and focus. This study aims to analyze the effect of the dominant orange color in the Shopee application interface on user focus levels. A quantitative correlational approach was employed, involving 32 respondents who completed a Likert-scale questionnaire, with data analyzed using SPSS. The results show a strong and statistically significant correlation between color perception and user focus (r = 0.681; p < 0.05), indicating that better evaluation of the dominant color is associated with higher focus levels. Furthermore, the coefficient of determination (R² = 0.464) reveals that nearly half of the variation in user focus is explained by color alone. These findings confirm that the dominant orange color enhances attention, supports sustained concentration, and improves navigation efficiency. This study highlights color as a key functional factor in optimizing user focus in mobile commerce applications.
Pemanfaatan Fingerprint Multi-Hash Ensemble dengan Mekanisme Veto untuk Mitigasi Fraud AI Berbasis Blockchain nawirah_athqiyah; Ali Ibrahim; Fathoni; Mgs Afriyan Firdaus
JURIKOM (Jurnal Riset Komputer) Vol. 13 No. 5 (2026): Oktober 2026
Publisher : Universitas Budi Darma

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30865/jurikom.v15i5.9732

Abstract

Proliferasi gambar hasil generasi AI seperti DALL-E, Midjourney, dan Stable Diffusion telah menciptakan ancaman serius terhadap autentikasi visual di berbagai platform digital. Kemampuan AI generatif dalam menghasilkan gambar fotorealistik membutuhkan pendekatan verifikasi yang robust dan efisien untuk memitigasi penyalahgunaan konten visual. Penelitian ini mengusulkan framework Multi-Hash Ensemble yang mengintegrasikan dua pendekatan fingerprinting, yaitu Full Image Hash Ensemble dan Hybrid Fingerprint dengan ROI Extraction, yang diperkuat oleh Veto Mechanism berbasis gradient signature, Local Binary Pattern (LBP), distribusi warna, dan shape descriptor, serta arsitektur blockchain on-chain/off-chain untuk menjamin keimutabelan data. Evaluasi dilakukan pada dataset 200 citra dari kategori face dan nature dengan 22.847 perbandingan melalui tiga skenario pengujian. Metode 1 (Full Image Hash Ensemble) menghasilkan True Acceptance Rate (TAR) 94,15%, True Rejection Rate (TRR) 98,97%, akurasi 98,43%, F1-Score 93,03%, dan Equal Error Rate (EER) 3,44%. Metode 2 (Hybrid Fingerprint) mencapai akurasi 96,71%, FAR terendah 0,68%, dan FAR nol pada mode HYBRID_ENSEMBLE untuk seluruh 3.486 pasangan imposter. Analisis sensitivitas threshold pada Metode 2 menunjukkan titik optimal pada 0,22 dengan F1-Score 85,18% dan akurasi 96,92%. Kedua metode menunjukkan ketahanan sempurna terhadap kompresi JPEG dan resize, namun memiliki kelemahan pada serangan Crop 10%. Dari aspek efisiensi blockchain, arsitektur hibrida on-chain/off-chain berhasil mereduksi data penyimpanan dari citra 2–5 MB menjadi hanya 64 byte per registrasi (rasio kompresi >99,99%), dengan estimasi biaya Gas Fee < $0,01 USD per citra pada jaringan Layer-2 (Polygon dan Arbitrum). Keimutabelan data off-chain dijamin melalui Content Hash SHA-256 yang dikunci pada blockchain, sehingga manipulasi data dapat dideteksi secara otomatis. Temuan ini menunjukkan bahwa framework yang diusulkan memberikan keseimbangan optimal antara akurasi deteksi, efisiensi penyimpanan, dan biaya operasional untuk verifikasi integritas citra digital skala besar.
Analisis Sentimen Ulasan Pengguna Alfagift di Google Play Store Menggunakan SVM dan Random Forest silvia nabila putri; Fahren Affandi; Ali Ibrahim; Fathoni
Buffer Informatika Vol. 12 No. 2 (2026): Buffer Informatika
Publisher : Department of Informatics Engineering, Faculty of Computer Science, University of Kuningan, Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.25134/buffer.v12i2.572

Abstract

Abstrak Meskipun Alfagift telah muncul sebagai platform belanja daring yang terkemuka di Indonesia, masih terdapat celah signifikan dalam studi sistematis mengenai kepuasan pengguna sebagaimana tercermin dalam ulasan aplikasi. Penelitian ini menjawab celah tersebut dengan mengevaluasi efektivitas algoritma Support Vector Machine (SVM) dan Random Forest dalam mengklasifikasikan sentimen pengguna ke dalam kategori positif, netral, dan negatif. Dengan memanfaatkan dataset sebanyak 2.000 ulasan yang diambil dari Google Play Store (rentang tahun 2025–2026), studi ini melibatkan pra-pemrosesan teks yang ketat, termasuk stemming dan vektorisasi TF-IDF. Untuk memitigasi tantangan ketidakseimbangan kelas, kedua model dioptimalkan menggunakan parameter class_weight='balanced'. Hasil eksperimen mengungkapkan bahwa SVM lebih mahir dalam memprediksi sentimen positif dengan akurasi 77,3%, sementara Random Forest menunjukkan presisi yang lebih tinggi dalam mendeteksi ketidakpuasan pengguna (sentimen negatif) dengan akurasi 88,6%. Pada akhirnya, studi ini berkontribusi pada literatur metodologi klasifikasi dan berfungsi sebagai alat evaluasi strategis bagi pengembang untuk meningkatkan kualitas layanan berdasarkan umpan balik pengguna. Kata Kunci: Alfagift; Analisis Sentimen; Random Forest; Support Vector Machine; Ulasan Pengguna;
Classification of Conflict Escalation and Destabilization in the Asia-Pacific Using Machine Learning Zwesty Quatra; Sella Juniastia Marsya Saputri; Imam Akbar; Fathoni
Buffer Informatika Vol. 12 No. 2 (2026): Buffer Informatika
Publisher : Department of Informatics Engineering, Faculty of Computer Science, University of Kuningan, Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.25134/buffer.v12i2.577

Abstract

The escalation of conflicts in the Asia-Pacific region risks triggering regional destabilization, yet evidence-based approaches to conflict prevention remain limited. This study aims to develop a machine learning framework to analyze the level of conflict escalation and support peacebuilding. Using a quantitative comparative design, secondary ACLED data from 2010–2025 were analyzed through a workflow encompassing data preprocessing, feature engineering, handling class imbalance using SMOTE, and evaluating four algorithms: Random Forest, XGBoost, Linear SVM, and Multi-layer Perceptron. Model performance was measured using accuracy, precision, recall, weighted F1 score, and macro AUC. The results showed that Random Forest performed best (accuracy 91.4%; weighted F1 score 91.5%; macro AUC 0.94), followed by XGBoost (accuracy 89.9%; macro AUC 0.93). Both ensemble models excelled in distinguishing low and moderate escalation classes, but their performance declined in the high and critical categories due to limitations in the original sample. Feature importance analysis revealed event type and frequency as the most significant predictors. These findings confirm the effectiveness of tree-based methods for structured conflict classification and their contribution to the development of early warning systems for conflict prevention in the Asia-Pacific region. The proposed framework supports anticipatory decision-making, timely intervention, and more efficient resource allocation for policymakers and international organizations.
Integrasi K-Means dan Random Forest untuk Prediksi Loyalitas Terhadap Retensi Pelanggan dalam Sistem CRM Gerri Asa Saputra; M. Putra Willy Nailis; M. Thoriqul Fadli; Ali Ibrahim; Fathoni
Buffer Informatika Vol. 12 No. 2 (2026): Buffer Informatika
Publisher : Department of Informatics Engineering, Faculty of Computer Science, University of Kuningan, Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.25134/buffer.v12i2.579

Abstract

Mempertahankan basis pelanggan adalah aspek penting dari Customer Relationship Management (CRM) pada era bisnis modern, tetapi analisis data transaksi mentah masih menghadapi masalah ketiadaan label loyalitas yang eksplisit. Penelitian ini mengusulkan pengembangan model analitik berbasis machine learning hibrida dengan mengintegrasikan algoritma K-Means dan Random Forest untuk mensegmentasi dan memprediksi loyalitas pelanggan secara objektif dan terukur. Studi kasus dilakukan menggunakan dataset publik transaksi ritel daring (online retail). Proses rekayasa fitur dilakukan menggunakan metode analisis Recency, Frequency, Monetary (RFM). Pendekatan unsupervised learning melalui K-Means diterapkan pada atribut Frequency dan Monetary yang telah ditransformasi untuk melabeli pelanggan ke dalam dua klaster utama, yakni loyal dan churn. Selanjutnya, Random Forest digunakan untuk memprediksi label pelanggan tersebut dengan mengisolasi variabel Recency sebagai prediktor tunggal guna mencegah kebocoran data (data leakage). Hasil implementasi dengan sampel data uji menunjukkan bahwa sistem ini mampu mengklasifikasikan pelanggan dengan akurasi keseluruhan sebesar 69,24%, dimana matriks konfusi mencatatkan identifikasi tepat sasaran sebanyak 241 pelanggan loyal dan 360 pelanggan churn. Temuan ini membuktikan bahwa integrasi K-Means dan Random Forest efektif sebagai alat prediksi loyalitas dan dapat dijadikan dasar perencanaan strategi pemasaran yang ditargetkan. Metode hibrida ini memberikan keunggulan dibandingkan analisis konvensional, yang biasanya bergantung pada asumsi klasifikasi rule-based tanpa mekanisme pelabelan algoritme yang jelas. Pendekatan ini tidak hanya menekan masalah ketiadaan variabel target, tetapi juga memberi perusahaan dasar analitis yang lebih kuat untuk menilai dan membuat keputusan tentang program retensi pelanggan.
Co-Authors Ali Ibrahim Ali Ibrahim Alifa Putri Shahabiyah Aliya Faiza Andin Sabilla Janna Badia Inaya Sazrade Fahren Affandi Fathimah Fadiyah Salimah Gerri Asa Saputra haniyah faizah Hendra Wijaya Imam Akbar Kevin Agustin Purba Purba M. Putra Willy Nailis M. Thoriqul Fadli Mgs Afriyan Firdaus Mira Afrina Muhammad Azyumardi Azra Muhammad Bayu Samudra Naek Parulian Hutagalung nawirah_athqiyah Oktavio Theonady Sella Juniastia Marsya Saputri Septi Mulia Putri silvia nabila putri Syalwa Salsabillah S Tamara Juliyanti Umar Rahman Zidan Zakirah Sabrina Putri Pasha Zwesty Quatra