Klasifikasi Berita Berdasarkan Kategori Menggunakan Convolutional Neural Network dengan IndoBERT
DOI:
https://doi.org/10.24843/JNATIA.2025.v03.i04.p20Keywords:
IndoBert, Convolutional Neural Network, Text Classification , Indonesian News, Contextual EmbeddingAbstract
The advancement of technology information has led to a significant increased the volume of digital news, that makes needs for automatic news classification. This study aims to design a model capable of caterogizing Indonesian language news articles into six predefined categories, such as News, Money, Bola, Health, Tekno, and Tren. To achieve this goal, the method used combines IndoBERT as the embedding technique with Convolutional Neural Network (CNN) as the classification algorithm. The dataset consists of 3.000 news articles collected from Kompas.com and is divided into training data and testing data using four different data split ratios: 60:40, 70:30 80:20, and 90:10 . The evaluation results show that the best performance was achieved using the 80:20 ratio, where the model reached an accuracy of 91%, along with high precision, recall, and F-1 Score These result prove that the combination of IndoBERT and CNN is effective for the automatic classification of Indonesian new texts.
Downloads
Published
Issue
Section
License
Copyright (c) 2025 Jonathan Federico Tantoro, I Dewa Made Bayu Atmaja Darmawan (Author)

This work is licensed under a Creative Commons Attribution 4.0 International License.