Comparative Analysis of RandAugment and AutoAugment on NASNetMobile for Online Gambling Image Classification in a Google Chrome Extension
Jurnal Informatika: Jurnal Pengembangan IT · 04.11.2026 · 11 (3) · 959-970
DOI: 10.30591/jpit.v11i3.54 ↗Kurz gefasst
Fachartikel in der Zeitschrift Jurnal Informatika: Jurnal Pengembangan IT, online erschienen am 04.11.2026. Verfasst von Fadhil Yuka Sahistya und 1 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
The rapid growth of internet technology has been accompanied by the widespread circulation of online gambling content across digital platforms, which continues to evolve and evade manual moderation, making automatic detection increasingly necessary. This study proposes an online gambling image classification system using NASNetMobile combined with the Random Augmentation (Rand Augment) technique. The model utilizes NASNetMobile pre-trained on ImageNet as a backbone, with Rand Augment applied during training to increase data diversity and improve generalization. Images are resized to 64x64 pixels with preprocessing according to NASNet standards. The classification layer consists of a Dense Layer with ReLu, Dropout, and Sigmoid output for binary classification. The model is evaluated using five random seeds and compared against AutoAugment and a non-augmented baseline. Results show that Rand Augment achieves the highest average accuracy of 85.03% and highest test accuracy of 87.42%, outperforming Auto Augment (81.26%) and baseline (82.85%), with average precision of 0.8529, recall of 0.8503, and F1-Score of 0.8501. Its standard deviation of 0.0144 indicates stable results despite slightly greater variance than the other methods. The system is implemented as a real-time Google Chrome extension using TensorFlow.js, enabling automatic detection and blurring of gambling content in the browser. These findings confirm that NASNetMobile combined with RandAugment offers an effective, lightweight, and efficient solution for browser-based classification.
Zitieren
APA
Sahistya, F. Y., & Naufal, M. (2026). Comparative Analysis of RandAugment and AutoAugment on NASNetMobile for Online Gambling Image Classification in a Google Chrome Extension. Jurnal Informatika: Jurnal Pengembangan IT, 11(3), 959-970. https://doi.org/10.30591/jpit.v11i3.54
Vancouver
Sahistya FY, Naufal M. Comparative Analysis of RandAugment and AutoAugment on NASNetMobile for Online Gambling Image Classification in a Google Chrome Extension. Jurnal Informatika: Jurnal Pengembangan IT. 2026;11(3):959-970. doi:10.30591/jpit.v11i3.54
BibTeX
@article{Sahistya2026,
author = {Sahistya, Fadhil Yuka and Naufal, Muhammad},
title = {Comparative Analysis of RandAugment and AutoAugment on NASNetMobile for Online Gambling Image Classification in a Google Chrome Extension},
journal = {Jurnal Informatika: Jurnal Pengembangan IT},
year = {2026},
volume = {11},
number = {3},
pages = {959--970},
doi = {10.30591/jpit.v11i3.54}
}
EUVIMED