ExaAEC is a multi-label Arabic emotion corpus that contains about 20,000 tweets classified as "neutral", "joy", "love", "anticipation", "acceptance", "surprise", "sadness", "fear", "anger" and "disgust". Each instance in the corpus contains the text and one or two most highlighted emotions that the writer wanted to express.
The text is separated by comma (',') from labels.
To have a vision on the general schema of the corpus, a tiny version of the corpus containing 100 samples has been provided named sample.csv
If you would like to access the whole corpus, please contact via [email protected] with subject: ExaAEC Request.
If you are using this dataset, please cite the initial paper
@INPROCEEDINGS{9721493,
author={Sarbazi-Azad, Saeed and Akbari, Ahmad and Khazeni, Mohsen},
booktitle={2021 11th International Conference on Computer Engineering and Knowledge (ICCKE)},
title={ExaAEC: A New Multi-label Emotion Classification Corpus in Arabic Tweets},
year={2021},
volume={},
number={},
pages={465-470},
doi={10.1109/ICCKE54056.2021.9721493}}