OIE2016 Dataset

OIE2016 is the first large-scale OpenIE benchmark. It is created by automatic conversion from QA-SRL [He et al., 2015], a semantic role labeling dataset. The sentences are from news (e.g., WSJ) and encyclopedia (e.g., WIKI) domains. Since there are no restrictions on the elements of OpenIE extractions, partial-matching criteria instead of exact-matching is typically used. Hence, the evaluation script can tolerate the extractions that are slightly different from the gold annotation.

Source: https://arxiv.org/pdf/2205.11725.pdf (section 3.1)

OIE2016 是首个大规模开放信息抽取（OpenIE）基准数据集。该数据集由自动转换自 QA-SRL（He 等人，2015）这一语义角色标注数据集所创建。数据句集来源于新闻（如 WSJ）和百科（如 WIKI）领域。鉴于开放信息抽取元素无特定限制，故通常采用部分匹配标准而非精确匹配标准。因此，评估脚本能够容忍与金标准标注略有差异的抽取结果。