CaRB

该数据集是目前最大的众包OIE数据集，提供了1,282个句子以及对应的二元组。这些黄金二元组以（主体，关系，对象）的格式呈现，并采用基于词级别的匹配方式进行评估。该数据集规模包含1,282个句子，任务是对开放式信息提取。

This is currently the largest crowdsourced Open Information Extraction (OIE) dataset, providing 1,282 sentences and their corresponding binary tuples. These golden binary tuples follow the format of (subject, relation, object) and are evaluated via word-level matching. With 1,282 sentences in total, this dataset is dedicated to the task of open information extraction.