SarcNet

URL: https://github.com/yuetanbupt/SarcNet

Description

SarcNet is a multilingual and multimodal sarcasm detection dataset that includes text-image pairs. The dataset contains 3,335 image-text pair samples in English and Chinese, with annotations for sarcasm in visual, textual, and multimodal data, totaling over 10,000 labeled instances. The dataset is designed to allow a more nuanced assessment of unimodal and multimodal models for sarcasm detection.

Dataset

• Size: 3,335 image-text pair samples
• Languages: English and Chinese
• Annotations: Over 10,000 labeled instances
  • Text annotations
  • Image annotations
  • Multimodal annotations

You can download the dataset from here.

Results

No specific results or performance metrics are provided in the repository.

Implementation & Code

The repository contains the dataset and guidelines for its usage in multimodal sarcasm detection experiments.