Benchmarking the Learnability of Aspect-Based Sentiment Analysis on Indonesian Protest Tweets
International Journal of Engineering, Science and Information Technology · 04.11.2026 · 6 (4) · 294-303
DOI: 10.52088/ijesty.v6i4.2006 ↗Kurz gefasst
Fachartikel in der Zeitschrift International Journal of Engineering, Science and Information Technology, online erschienen am 04.11.2026. Verfasst von Putri Rizqiyah und 2 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
Aspect-based sentiment analysis (ABSA) on Indonesian social media has steadily improved in accuracy, yet the field still lacks evidence on which ABSA subtasks are learnable on imbalanced, event-driven text and how much of a reported score reflects the protocol rather than the model. This study benchmarks relevance, aspect, and aspect-level polarity classification on 7,131 tweets from the August 2025 Indonesian demonstrations, against a documented gold standard of 6,075 relevant tweets. Majority-class prediction, four lexical models, a fine-tuned IndoBERT, and zero-shot GPT-4o-mini are compared under identical five-fold stratified cross-validation, with macro-F1 as the primary metric, with 95% paired-bootstrap and McNemar tests on identical folds, and with a repeated benchmark under near-duplicate-grouped and day-held-out splits. Headline accuracy is misleading for relevance: models reach 85-89% against an 85.19% majority baseline, yet lexical models detect only one in ten off-topic tweets (recall 0.099) and the fine-tuned encoder no more than half (0.495). Aspect classification is the most learnable subtask, at 0.772 macro-F1 for the fine-tuned encoder, and zero-shot GPT-4o-mini reaches 0.808 on the 78.9% of records it chose to label, against 0.711 when its omissions are counted as errors. Positive polarity is the bottleneck for every family: lexical models reach 0.306 F1 with 0.190 recall, and although contextual and zero-shot models recover positive recall to roughly 0.45, F1 stays in the 0.51-0.57 range on the records they label. Sentiment toward the people aspect is the hardest cell at 57.5% accuracy, against 86.6% for the government. The zero-shot comparator labels 84.2% of the relevant tweets for polarity, and its errors are systematic: 68.2% over-predict the negative class, conflating event valence with evaluative polarity. The study contributes a task-level, coverage-aware benchmark for Indonesian ABSA under class imbalance
Zitieren
APA
Rizqiyah, P., Fauzi, A. R., & Sazkia, A. C. (2026). Benchmarking the Learnability of Aspect-Based Sentiment Analysis on Indonesian Protest Tweets. International Journal of Engineering, Science and Information Technology, 6(4), 294-303. https://doi.org/10.52088/ijesty.v6i4.2006
Vancouver
Rizqiyah P, Fauzi AR, Sazkia AC. Benchmarking the Learnability of Aspect-Based Sentiment Analysis on Indonesian Protest Tweets. International Journal of Engineering, Science and Information Technology. 2026;6(4):294-303. doi:10.52088/ijesty.v6i4.2006
BibTeX
@article{Rizqiyah2026,
author = {Rizqiyah, Putri and Fauzi, Ahmad Ridwan and Sazkia, Adelia Chitra},
title = {Benchmarking the Learnability of Aspect-Based Sentiment Analysis on Indonesian Protest Tweets},
journal = {International Journal of Engineering, Science and Information Technology},
year = {2026},
volume = {6},
number = {4},
pages = {294--303},
doi = {10.52088/ijesty.v6i4.2006}
}
EUVIMED