README.md

metadata

license: mit
configs:
  - config_name: en
    data_files:
      - split: train
        path: en/train-*
      - split: validation
        path: en/validation-*
      - split: test
        path: en/test-*
  - config_name: zh
    data_files:
      - split: train
        path: zh/train-*
      - split: validation
        path: zh/validation-*
      - split: test
        path: zh/test-*
dataset_info:
  - config_name: en
    features:
      - name: id
        dtype: string
      - name: image
        dtype: image
      - name: text
        dtype: string
      - name: text_label
        dtype:
          class_label:
            names:
              '0': non-sarcastic
              '1': sarcastic
      - name: image_label
        dtype:
          class_label:
            names:
              '0': non-sarcastic
              '1': sarcastic
      - name: multi_label
        dtype:
          class_label:
            names:
              '0': non-sarcastic
              '1': sarcastic
      - name: language
        dtype: string
    splits:
      - name: train
        num_bytes: 56554731
        num_examples: 758
      - name: validation
        num_bytes: 17782457
        num_examples: 247
      - name: test
        num_bytes: 19742839
        num_examples: 267
    download_size: 93869028
    dataset_size: 94080027
  - config_name: zh
    features:
      - name: id
        dtype: string
      - name: image
        dtype: image
      - name: text
        dtype: string
      - name: text_label
        dtype:
          class_label:
            names:
              '0': non-sarcastic
              '1': sarcastic
      - name: image_label
        dtype:
          class_label:
            names:
              '0': non-sarcastic
              '1': sarcastic
      - name: multi_label
        dtype:
          class_label:
            names:
              '0': non-sarcastic
              '1': sarcastic
      - name: language
        dtype: string
    splits:
      - name: train
        num_bytes: 166326882
        num_examples: 1240
      - name: validation
        num_bytes: 65125110
        num_examples: 424
      - name: test
        num_bytes: 58952182
        num_examples: 399
    download_size: 293031710
    dataset_size: 290404174
language:
  - en
  - zh
tags:
  - sarcasm
  - Multimodal
  - multilingual

SarcNet: A Multilingual Multimodal Sarcasm Detection Dataset

SarcNet is a novel benchmark for multilingual and multimodal sarcasm detection, introduced at LREC-COLING 2024. It addresses the limitations of single-language datasets by providing 3,335 image-text pair samples in both English and Chinese.

Dataset Summary

In contrast to traditional datasets that use a single unified label, SarcNet employs a separated annotation schema. Each sample is distinctly labeled across three modalities, providing over 10,000 labels in total:

• Text Label: Sarcasm detected in the text alone.
• Image Label: Sarcasm detected in the image alone.
• Multimodal Label: Sarcasm detected through the combination of both modalities.

Data Splits

The dataset follows an official split of 60% Train, 20% Val, and 20% Test.

Language	Train	Validation	Test	Total
Chinese	1,242	424	399	2,065
English	756	247	267	1,270
Total	1,998	671	666	3,335

Dataset Features

The columns in this reformatted version are:

• id: The unique identifier (original image path).
• image: The decoded image file.
• text: The associated textual statement.
• text_label: ClassLabel (0: non-sarcastic, 1: sarcastic).
• image_label: ClassLabel (0: non-sarcastic, 1: sarcastic).
• multi_label: ClassLabel (0: non-sarcastic, 1: sarcastic).
• language: The detected language code (en or zh).

Usage

Because this dataset is configured with subsets, you can load your desired language directly:

from datasets import load_dataset

# Load English only
dataset_en = load_dataset("alita9/sarcnet", "en")

# Load Chinese only
dataset_zh = load_dataset("alita9/sarcnet", "zh")

Citation

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

@inproceedings{yue-etal-2024-sarcnet, title = "{S}arc{N}et: A Multilingual Multimodal Sarcasm Detection Dataset", author = "Yue, Tan and Shi, Xuzhao and Mao, Rui and Hu, Zonghai and Cambria, Erik", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", year = "2024", address = "Torino, Italia", pages = "14325--14335", publisher = "ELRA Language Resource Association", url = "https://aclanthology.org/2024.lrec-main.1248" }