license: cc-by-3.0 | |
RoBERTA-base finetuned on our Dissonance Dataset, collected from annotating tweets for within-person dissonance, as described in our paper [Transfer and Active Learning for Dissonance Detection: Addressing the Rare Class Challenge](https://arxiv.org/abs/2305.02459). | |
## Dataset Annotation details | |
Tweets were parsed into discourse units, and marked as Belief (Thought or Action) or Other, and pairs of beliefs within the same tweet were relayed to annotators for Dissonance annotation. | |
![annotation process](./annotation_format/annotation_process.jpg) | |
The annotations were conducted on a sheet in the following **dissonance-first** format. | |
![annotation format](./annotation_format/annotation_format.png) | |
The annotators used the following flowchart as a more detailed guide to determining the Dissonance, Consonance and Neither/Other classes: | |
![annotation guidelines](./annotation_format/annotation_guidelines.jpg) | |