ViTASA: New benchmark and methods for Vietnamese targeted aspect sentiment analysis for multiple textual domains

Published: 01 Jan 2025, Last Modified: 07 May 2025Comput. Speech Lang. 2025EveryoneRevisionsBibTeXCC BY-SA 4.0
Abstract: Highlights•Creating multi-domain dataset with 500k+ target-aspect pairs from social comments.•Proposing transformer-based approach achieving SOTA on Vietnamese TASA task.•Evaluating Vietnamese multi-task sentiment analysis including TBSA and ABSA tasks.•Investigating challenges and characteristics in Vietnamese sentiment analysis tasks.
Loading