EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Beyond Opinion Terms: A Hybrid Token-Span Architecture for Atomic Opinion Quote Extraction in Arabic

Bayan Aldashnan, Abdulrahman Alothaim, Ahmed Alsanad

IEEE Access · 2026 · 14 · 91176-91186

DOI: 10.1109/access.2026.3703127 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Access, 2026. Verfasst von Bayan Aldashnan und 2 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Aspect-based sentiment analysis identifies fine-grained opinions in user reviews, but most systems simply predict sentiment labels without revealing the text spans that justify those labels. Existing approaches extract minimal opinion terms, which cannot verify predictions, explain reviewer attitudes, or support downstream applications such as aspect classification. Addressing this, we introduce atomic opinion quote extraction, a formulation that extracts complete, minimal, self-contained evaluative spans exactly as they appear in the original text, structured as (aspect category, sentiment polarity, opinion quote) triplets. We created a large-scale dataset of 14,783 real-world restaurant reviews written in Arabic, spanning diverse dialects and annotated with 63,620 opinion triplets covering 29 aspect categories. This represents a benchmark for atomic opinion quote extraction in Arabic. We further propose a hybrid token–span architecture that combines sequence labeling with a boundary-aware span classifier to extract the atomic opinion quotes with precise boundary detection. Experimental results show that, on this challenging extraction task, the proposed model achieves 65.34% exact-match F1 and 81.67% F1 at a 50% overlap threshold, consistently outperforming strong token- and span-based baselines with statistically significant improvements. Combined, this task, dataset, and model advance Arabic sentiment analysis beyond mere label prediction toward explainable, fine-grained, and evidence-based opinion understanding, with broad applicability to downstream Arabic natural language processing tasks.

Zitieren

APA

Aldashnan, B., Alothaim, A., & Alsanad, A. (2026). Beyond Opinion Terms: A Hybrid Token-Span Architecture for Atomic Opinion Quote Extraction in Arabic. IEEE Access, 14, 91176-91186. https://doi.org/10.1109/access.2026.3703127

Vancouver

Aldashnan B, Alothaim A, Alsanad A. Beyond Opinion Terms: A Hybrid Token-Span Architecture for Atomic Opinion Quote Extraction in Arabic. IEEE Access. 2026;14:91176-91186. doi:10.1109/access.2026.3703127

BibTeX
@article{Aldashnan2026,
  author = {Aldashnan, Bayan and Alothaim, Abdulrahman and Alsanad, Ahmed},
  title = {Beyond Opinion Terms: A Hybrid Token-Span Architecture for Atomic Opinion Quote Extraction in Arabic},
  journal = {IEEE Access},
  year = {2026},
  volume = {14},
  pages = {91176--91186},
  doi = {10.1109/access.2026.3703127}
}
Kostenlos starten →