Large Language Models as Automatic Annotators and Annotation Adjudicators for Fine-Grained Opinion Analysis 事件

Large Language Models as Automatic Annotators and Annotation Adjudicators for Fine-Grained Opinion Analysis arXiv:2601.16800v3 Announce Type: replace Abstract: Fine-grained opinion analysis of text provides a detailed understanding of expressed sentiments, including the addressed entity. Although this level of detail is valuable, annotating opinions in datasets for model training requires considerable human effort and substantial cost, especially across diverse domains and real-world applicatio