xSID

该数据集为跨语言的槽位和意图检测提供了一个基准，包含了13种语言的平行标注数据。此外，该数据集总共包含了18个意图标签用于分类。在规模上，每种语言都有10000个训练句子、300个验证句子和300个测试句子。该数据集的任务是意图分类。

This dataset serves as a benchmark for cross-lingual slot and intent detection, featuring parallel annotated data across 13 languages. It encompasses 18 intent labels for classification tasks. Regarding dataset scale, each language has 10,000 training sentences, 300 validation sentences, and 300 test sentences. The primary task of this dataset is intent classification.