Automated News Categorisation Using Transformer-Based Deep Learning Models: A Step Toward Intelligent Media Classification
Pertanika Journal of Science and Technology · 30.10.2026 · 34 (5)
DOI: 10.47836/pjst.34.5.21 ↗Kurz gefasst
Fachartikel in der Zeitschrift Pertanika Journal of Science and Technology, online erschienen am 30.10.2026. Verfasst von B. Rupa Devi und 5 weiteren Autor:innen.
Abstract (Originalsprache)
The rapid growth of online news content has increased the need for automated methods to achieve accurate and scalable news categorisation. Manual classification is labour-intensive and difficult to maintain when processing large volumes of heterogeneous news articles. To address this challenge, this study proposes a Transformer-based deep learning framework for automated news classification. The proposed architecture incorporates positional encoding and multi-head self-attention to capture contextual dependencies among words, while dropout regularisation is employed to reduce overfitting. In addition, domain-specific fine-tuning is applied to adapt the pre-trained Transformer representations to the characteristics of news classification. The proposed model is evaluated using the HuffPost News Category Dataset, which contains approximately 210,000 news articles distributed across 42 categories. Experimental results demonstrate that the proposed framework achieves 96.78% accuracy on the evaluation set. For comparison, a Multinomial Naïve Bayes (MNB) classifier achieves an accuracy of 84.52%. These results indicate that the Transformer-based framework provides improved classification performance compared with the selected conventional baseline. However, further evaluation using additional performance metrics, statistical significance testing, and comparisons with contemporary deep learning and Transformer-based baselines is necessary to establish the robustness and generalisability of the proposed approach, Long Short-Term Memory (LSTM) (89.63%), and BERT (92.14%). Ablation data also confirm the usefulness and need for the integrated architectural elements. The proposed design successfully handles the complexity of large-scale, multi-class news classification, representing a crucial step forward in automated content analysis.
Zitieren
APA
Devi, B. R., Gayatri, M., Lingam, R., Balaji, P., Beatrice, S., & Athinarayanan, S. (2026). Automated News Categorisation Using Transformer-Based Deep Learning Models: A Step Toward Intelligent Media Classification. Pertanika Journal of Science and Technology, 34(5). https://doi.org/10.47836/pjst.34.5.21
Vancouver
Devi BR, Gayatri M, Lingam R, Balaji P, Beatrice S, Athinarayanan S. Automated News Categorisation Using Transformer-Based Deep Learning Models: A Step Toward Intelligent Media Classification. Pertanika Journal of Science and Technology. 2026;34(5). doi:10.47836/pjst.34.5.21
BibTeX
@article{Devi2026,
author = {Devi, B. Rupa and Gayatri, Malyala and Lingam, Rajesh and Balaji, Penubaka and Beatrice, S. and Athinarayanan, S.},
title = {Automated News Categorisation Using Transformer-Based Deep Learning Models: A Step Toward Intelligent Media Classification},
journal = {Pertanika Journal of Science and Technology},
year = {2026},
volume = {34},
number = {5},
doi = {10.47836/pjst.34.5.21}
}
EUVIMED