shijli/semeval2022-task5

Dataset Card for Multimedia Automatic Misogyny Identification Task of Semeval2022

Dataset Summary

The datasets (10000 + 1000 memes about misogyny detection) are exclusively reserved for the participants of SemEval-2022 Task 5 and are not to be used freely. The data may be distributed upon request and for academic purposes only. To request the datasets, please fill out the following form: https://forms.gle/AGWMiGicBHiQx4q98 After submitting the required info, participants will have a link to a folder containing the datasets in a zip format (trial, training and development) and the password to uncompress the files.

How to Use

import datasets

datasets.load_dataset(
    "shijli/semeval2022-task5", 
    training_dir="/path/to/training_images/",
    test_dir="/path/to/test_images/",
)

File Structure

Please ensure your dataset is organised in the following structure:

training_dir
├── 1.jpg
├── 2.jpg
├── ...
├── training.csv

test_dir
├── 1.jpg
├── 2.jpg
├── ...
├── test.csv
├── test_labels.txt

Citation Information

If you use the MAMI dataset, please cite the following paper:

@inproceedings{fersini2022,
    title={{SemEval-2022 Task 5}: Multimedia Automatic Misogyny Identification},
    author={Elisabetta Fersini, Francesca Gasparini, Giulia Rizzi, Aurora Saibene, Berta Chulvi, Paolo Rosso, Alyssa Lees, Jeffrey Sorensen},
    booktitle={Proceedings of the 16th International Workshop on Semantic Evaluation (SemEval-2022)},
    year={2022},
    publisher={Association for Computational Linguistics}
}