Stance Classification Towards Political Figures on Blog Writing
there is two dataset model that was built for stance detection research.
First research by Cakra W in 2016. He was using three labels for building the dataset. three label is
- For - the text that is created by author is support the target in an event
- Against - the text that is created by author is oppose the target in an event
- Unknown - the text that is created by author, we don't know the text is support or oppose the target the dataset is about three target on same events
Second research by Rini J in 2018. She was using two labels for building the dataset. This dataset is combined of cakra dataset and new dataset. Rini and team decided to re-annotated cakra dataset using two labels. Two label is only 'For' and 'Againts'. the dataset is about five target and every target have 1 different event.
On experiments we use stopwords by Tala [1], sentiment lexicon by Vania et al.[2], and word2vec that is built from newspaper source, you can see in other work in https://drive.google.com/open?id=1uzWUGVXDlqyMJ7AL9deQBZTDTOTLDJCn.
This dataset and the other resource can be used for free, but if you want to publish paper/publication using this dataset, please cite this publication:
R. Jannati, R. Mahendra, C. W. Wardhana and M. Adriani, "Stance Classification Towards Political Figures on Blog Writing," 2018 International Conference on Asian Language Processing (IALP), Bandung, Indonesia, 2018, pp. 96-101, doi: 10.1109/IALP.2018.8629144. IEEE Xplore https://ieeexplore.ieee.org/document/8629144
or you can copy this BibTex
@INPROCEEDINGS{8629144,
author={R. {Jannati} and R. {Mahendra} and C. W. {Wardhana} and M. {Adriani}},
booktitle={2018 International Conference on Asian Language Processing (IALP)},
title={Stance Classification Towards Political Figures on Blog Writing},
year={2018},
volume={},
number={},
pages={96-101},
}
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.
