“…Existing work on agreement analysis is aimed at (1) identifying language that indicates disagreement (e.g., Niculae and Danescu-Niculescu-Mizil, 2016;Wojatzki et al, 2018;Fischer et al, 2022), (2) leveraging stylistic choices like sarcasm for detecting disagreement (Ghosh et al, 2021), (3) finding stance and target pairs, followed by the traditional stance classification (e.g., Chen et al, 2019;De Kock and Vlachos, 2021), and (4) mixing de-tailed opinion information using e.g., logic of evaluation (Draws et al, 2022). Recently, adding social role context to textual comments was shown to have a positive impact on the agreement analysis task (Luo et al, 2023), which indicates the usefulness of background information.…”