Article Per Year (5 Year)
p-Index From 2021 - 2026
3.792
P-Index
This Author published in this journals
All Journal Teknika Jurnal Ilmu Teknik dan Komputer BERNAS: Jurnal Pengabdian Kepada Masyarakat Journal of Computer Networks, Architecture and High Performance Computing Jurnal Pengabdian Masyarakat (ABDIRA) Jurnal Informatika dan Komputer (JIK) Jurnal Sistem Informasi Mahakarya (JSIM) Jurnal Ilmiah Betrik : Besemah Teknologi Informasi dan Komputer
sri hartati
Universitas Mahakarya Asia
Religion Aerospace Engineering Agriculture, Biological Sciences & Forestry Humanities Automotive Engineering Chemical Engineering, Chemistry & Bioengineering Civil Engineering, Building, Construction & Architecture Computer Science & IT Control & Systems Engineering Dentistry Education Electrical & Electronics Engineering Engineering Environmental Science Industrial & Manufacturing Engineering Languange, Linguistic, Communication & Media Mathematics Neuroscience Nursing Public Health Social Sciences Veterinary Other

Published : 24 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 24 Documents
Search

 1   2   3 
Performance Analysis of an Offline Text Detection System Based on Edge AI A Case Study of DokuScan Pro Defi Pujianto; Kadarsih Kadarsih; Sri Hartati
Journal of Computer Networks, Architecture and High Performance Computing Vol. 8 No. 1 (2026): Articles Research Januari 2026
Publisher : Information Technology and Science (ITScience)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47709/cnahpc.v8i1.7854

Abstract

The growing use of mobile document scanning applications has increased the demand for text detection systems that can operate reliably in offline and on-device environments. Although Edge AI enables local inference without network dependency, system-level empirical evidence regarding its performance under real-world mobile usage conditions remains limited. This study presents a system-level evaluation of an offline Edge AI–based text detection system for mobile document scanning, using DokuScan Pro as a case study. The evaluation was conducted on 40 document images captured under varying lighting conditions, capture angles, and background characteristics. System performance was assessed using precision, recall, F1-score, and inference time to characterize on-device behavior rather than algorithmic novelty. Experimental results show that the system achieved a precision of 1.00, a recall of 0.975, and an F1-score of approximately 0.98, with an average inference time of 63.8 ms per image during fully offline execution on mobile devices. These results indicate stable system-level performance under real-world document scanning conditions with controlled computational overhead. This study provides empirical system-level insights into the feasibility and practical limitations of deploying Edge AI–based text detection in offline mobile document scanning applications, thereby complementing existing model-centric research with evidence from real-world, on-device evaluation.
Perbandingan Kinerja Isolation Forest Dan Local Outlier Factor (LOF) Dalam Deteksi Anomali Transaksi Digital Sri Hartati; Defi Pujianto; Kadarsih
BETRIK Vol. 17 No. 01 (2026): Jurnal Ilmiah BETRIK : Besemah Teknologi Informasi dan Komputer
Publisher : PPPM Institut Teknologi Pagar Alam

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.36050/5w76ab74

Abstract

The rapid growth of digital transactions has increased the risk of anomalous activities such as fraud, particularly in highly imbalanced datasets where fraudulent transactions are significantly fewer than normal transactions. This imbalance presents a major challenge in anomaly detection, as models tend to be biased toward the majority class. This study aims to compare the performance of Isolation Forest and Local Outlier Factor (LOF) algorithms in detecting anomalies in digital transaction data.The research adopts an experimental approach using the Credit Card Fraud Detection dataset, which consists of 284,807 transactions, including 492 fraudulent cases. Data preprocessing involves feature normalization using StandardScaler, followed by a stratified train-test split with a ratio of 70:30. Model evaluation is conducted using confusion matrix, precision, recall, and F1-score metrics.The results show that Isolation Forest outperforms LOF. Isolation Forest successfully detects 37 out of 148 fraudulent transactions with a precision of 0.2824, recall of 0.25, and F1-score of 0.2652. In contrast, LOF detects only 2 fraudulent transactions, with a precision of 0.0137, recall of 0.0135, and F1-score of 0.0136. These findings indicate that isolation-based approaches are more effective and robust than density-based methods in handling highly imbalanced datasets.
Explainable Predictive Analytics untuk Prediksi Pengunduran Diri Karyawan pada Data Human Resource Analytics Sri Hartati; Rusidi; Dodi Herryanto
BETRIK Vol. 17 No. 02 (2026): Jurnal Ilmiah BETRIK : Besemah Teknologi Informasi dan Komputer
Publisher : PPPM Institut Teknologi Pagar Alam

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.36050/gcx9tn49

Abstract

Digital transformation has encouraged organizations to adopt Human Resource Analytics and Artificial Intelligence to support data-driven decision-making, including employee attrition prediction. Although numerous predictive models have been developed, most of them still suffer from limited interpretability, making their predictions difficult for Human Resource practitioners to understand and utilize. This study aims to develop an Explainable Predictive Analytics model for employee attrition prediction by integrating Information Gain, Random Forest, RandomizedSearchCV, and SHapley Additive exPlanations (SHAP). The study employs the IBM HR Analytics Employee Attrition & Performance dataset consisting of 1,470 employee records. The research workflow includes data preprocessing, feature selection using Information Gain, Random Forest model development, hyperparameter optimization using RandomizedSearchCV, model evaluation using Accuracy, Precision, Recall, F1-Score, and ROC-AUC, followed by model interpretation through SHAP Summary Plot and SHAP Feature Importance. The experimental results indicate that the model achieved 82.54% Accuracy, 37.50% Precision, 12.68% Recall, 18.95% F1-Score, and a ROC-AUC of 0.7439. Feature selection results indicate that OverTime has the highest Information Gain value, while MonthlyIncome is identified as the most influential feature according to Random Forest Feature Importance. The main contribution of this study is the integration of Information Gain-based feature selection, Random Forest optimization, and SHAP-based Explainable Artificial Intelligence into a unified Explainable Predictive Analytics framework, providing a more transparent and interpretable predictive model to support decision-making in Human Resource Management
Оptimasi Mоdel Hybrid Randоm Fоrest dan Gradient Bооsting dalam Sentimen Ulasan Pengguna Shоpee Sri Hartati; Muhajir Arafat; Rusidi Rusidi
TEKNIKA Vol. 19 No. 3 (2025): Teknika September 2025
Publisher : Politeknik Negeri Sriwijaya

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.5281/zenodo.17043356

Abstract

Dalam era transfоrmasi digital, penggunaan e-cоmmerce telah melekat erat dalam keseharian masyarakat Indоnesia, dengan aplikasi Shоpee sebagai salah satu platfоrm terpоpuler. Ulasan pengguna yang melimpah menawarkan peluang besar untuk memahami persepsi kоnsumen, namun analisis sentimen pada teks berbahasa Indоnesia masih menghadapi tantangan karena sifatnya yang infоrmal dan tidak seimbang. Penelitian ini bertujuan mengоptimalkan klasifikasi sentimen ulasan Shоpee melalui pendekatan hybrid stacking yang menggabungkan Randоm Fоrest(RF) dan Gradient Bооsting (GB) dengan Lоgistic Regressiоn sebagai meta learner. Dataset ulasan Shоpee diprоses menggunakan teknik preprоcessing teks dan ekstraksi fitur TF IDF. Evaluasi perfоrma dilakukan dengan metrik akurasi, precisiоn, recall, dan F1-scоre. Hasil menunjukkan bahwa mоdel Hybrid Stacking memiliki kinerja paling unggul dengan akurasi 0,82, precisiоn 0,82, recall 0,64, dan F1-scоre 0,65. Sementara mоdel Randоm Fоrest mencatatkan akurasi 0,80, precisiоn 0,86, recall 0,59, dan F1-scоre 0,59. Adapun mоdel Gradient Bооsting memperоleh akurasi 0,77, precisiоn 0,62, recall 0,57, dan F1-scоre 0,55. Hasil ini menunjukkan bahwa pendekatan hybrid stacking mampu meningkatkan perfоrma klasifikasi dibandingkan mоdel individual. Temuan ini memberikan kоntribusi terhadap pengembangan sistem analisis sentimen yang lebih akurat. Temuan ini menegaskan pentingnya pendekatan hybrid dalam analisis sentimen e-cоmmerce berbahasa Indоnesia, dan memberikan nilai tambah bagi pembangunan sistem pendukung keputusan yang lebih fleksibel dan akurat.
Co-Authors A Latif Abdulloh Bella Tania Bina Aji Satrio Budi Kurniawan Defi Pujianto Destia Louriska Matheos Dian Meilantika Dian Sri Agustina Dodi Dodi Herryanto Erlita Sulistiati Fatimah Kesuma Astuti Haris Saputro Hendrayudi Hendrayudi Herwa Herwa Iftia Arianti Johansen Juherni Marsella Kadarsih Kadarsih Kadarsih M As'ad Maulana Muhajir Arafat Nagris Sunggariang Naldo Nia Andriani Nilam Permata Noor Safitri Mianti Nur Fadhilla Rahmat Riski Riska Widiana Rusidi Rusidi Rusidi Salamudin Salamudin salamudin salamudin Sarkowi Satria Novari Suryanto Suryanto Taufik Sobri Wisnumurti