ArSarcasm-v2

该数据集名为ArSarcasm-v2，是一个专门用于讽刺和情感分析的数据集，包含了标记有情感、讽刺以及方言的阿拉伯语推文。每条推文都被标注了情感（正面、中性、负面）、讽刺（是、否）以及方言（现代标准阿拉伯语、埃及方言、黎凡特方言、马格里布方言、海湾方言）。该数据集规模约为15,548条推文，其中12,000条用于训练，3,000条用于测试。其任务涵盖了情感分析、讽刺检测以及方言识别。

This dataset, named ArSarcasm-v2, is a specialized resource for sarcasm and sentiment analysis, consisting of Arabic tweets annotated with sentiment, sarcasm, and dialect information. Each tweet is labeled with sentiment categories (positive, neutral, negative), sarcasm status (yes, no), and dialect type (Modern Standard Arabic, Egyptian Arabic, Levantine Arabic, Maghrebi Arabic, Gulf Arabic). The dataset comprises approximately 15,548 tweets in total, with 12,000 samples allocated for training and 3,000 for testing. The supported tasks of this dataset include sentiment analysis, sarcasm detection, and dialect identification.