Datasets:

alita9
/

muse-sarcasm-explanation

license: mit
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
      - split: validation
        path: data/validation-*
      - split: test
        path: data/test-*
dataset_info:
  features:
    - name: id
      dtype: string
    - name: image
      dtype: image
    - name: text
      dtype: string
    - name: label
      dtype:
        class_label:
          names:
            '0': sarcastic
    - name: explanation
      dtype: string
    - name: is_ocr
      dtype: bool
  splits:
    - name: train
      num_bytes: 333867120
      num_examples: 2983
    - name: validation
      num_bytes: 19589769
      num_examples: 175
    - name: test
      num_bytes: 38841746
      num_examples: 352
  download_size: 390979603
  dataset_size: 392298635

MuSe: Multimodal Sarcasm Explanation (Reformatted)

This repository provides a Hugging Face-compatible version of the MuSe (MORE) dataset.

Dataset Description

Original Authors: LCS2-IIITD (Shweta Desai, et al.)
Original Source: GitHub Repository
Paper: Nice perfume. How long did you marinate in it? Multimodal Sarcasm Explanation (AAAI 2022)

Modifications in this version

To make the dataset easier to use with the datasets library, the following changes were made:

Unified Schema: Merged separate OCR and Non-OCR files into a single test split.
Metadata Flags: Added an is_ocr (boolean) column to distinguish between image types.
Image Integration: Converted image paths into a native Hugging Face Image feature for direct loading.
Labeling: Explicitly labeled all samples as "sarcastic" to match standard classification formats.

Citation

If you use this dataset in your research, please cite the original work:

@inproceedings{desai2022nice, title={Nice perfume. How long did you marinate in it? Multimodal Sarcasm Explanation}, author={Desai, Shweta and Pawar, Tanmay and Shah, Parth and Mittal, Akshat and Das, Amitava and Shah, Rajiv Ratn and others}, booktitle={Proceedings of the AAAI Conference on Artificial Intelligence}, year={2022} }

License

This dataset is distributed under the MIT License, as per the original source repository.