Building models for stance detection on different Datasets. We build linear SVM and logistic regression models on the different datasets and test cross-dataset performance.
Datasets used:
- SemEval 2016 Task 6 A and B: http://alt.qcri.org/semeval2016/task6/
- Multi-Target Stance Detection: https://www.aclweb.org/anthology/E17-2088.pdf
- Constance: https://github.com/kennyjoseph/constance
Please note that due to Twitter's TOS, we cannot make the tweets available. Please download the datasets from the above links or contact us for access.