CAMR 2.0

该数据集名为CAMR 2.0，旨在评估中文抽象意义表示（AMR）解析系统的性能。它包含了训练集、开发集和测试集，以及一个用于评估泛化性能的跨领域盲测集。此外，该数据集还包括了概念标注和关系分类任务，模型性能将通过不同阶段的各类F1分数来评估。在规模上，数据集涵盖了31941个不同的概念，8443个不同的谓词以及142种不同的关系。该任务专注于中文AMR解析。

This dataset, named CAMR 2.0, is designed to evaluate the performance of Chinese Abstract Meaning Representation (AMR) parsing systems. It consists of training, development, and test splits, as well as a cross-domain blind test set for assessing generalization performance. Additionally, this dataset includes concept annotation and relation classification tasks, and model performance will be evaluated using various F1 scores across different stages. In terms of scale, the dataset covers 31,941 distinct concepts, 8,443 distinct predicates, and 142 distinct relation types. This task focuses on Chinese AMR parsing.