IndoRoBERTa-Base-PDBNews-Category

Evaluation

The suggested GDP News Indonesian Online classification model on Category Industry task yielded an 86.17% F1-score and an accuracy of 85.88% (Epoch=10, Batch Size=32)

Citation

• Plain Text: E. Fitria, A. Afnan and T. F. Abidin, "Detecting GDP Movement and Growth Rates from Indonesian Economic News Using IndoRoBERTa with Neural Sentence Selection," 2026 18th International Conference on Knowledge and Smart Technology (KST), Pattaya, Thailand, 2026, pp. 545-550, doi: 10.1109/KST67832.2026.11432333.
• BibTex: @INPROCEEDINGS{11432333, author={Fitria, Eka and Afnan, Afnan and Abidin, Taufik Fuadi}, booktitle={2026 18th International Conference on Knowledge and Smart Technology (KST)}, title={Detecting GDP Movement and Growth Rates from Indonesian Economic News Using IndoRoBERTa with Neural Sentence Selection}, year={2026}, volume={}, number={}, pages={545-550}, keywords={Economics;Accuracy;Economic indicators;Biological system modeling;Diversity reception;Transformers;Real-time systems;Macroeconomics;Monitoring;Tuning;Gross Domestic Product (GDP);online economic news;text classification;transformer models;IndoRoBERTa;neural sentence selection}, doi={10.1109/KST67832.2026.11432333}}

@INPROCEEDINGS{11432333, author={Fitria, Eka and Afnan, Afnan and Abidin, Taufik Fuadi}, booktitle={2026 18th International Conference on Knowledge and Smart Technology (KST)}, title={Detecting GDP Movement and Growth Rates from Indonesian Economic News Using IndoRoBERTa with Neural Sentence Selection}, year={2026}, volume={}, number={}, pages={545-550}, keywords={Economics;Accuracy;Economic indicators;Biological system modeling;Diversity reception;Transformers;Real-time systems;Macroeconomics;Monitoring;Tuning;Gross Domestic Product (GDP);online economic news;text classification;transformer models;IndoRoBERTa;neural sentence selection}, doi={10.1109/KST67832.2026.11432333}}

Acknowledgment

The authors express their gratitude to all internal and external sources that consistently provide support for our study, especially to Wilson Wongso, Steven Limcorn, Cahya Wirawan, Yogie O. Sihombing, M. J. Mubarok for their openly usable models on HuggingFace.