Wikipedia Event Coreference (WEC)

Wikipedia Event Coreference (WEC) 数据集是由以色列巴伊兰大学创建的大规模跨文档事件共指数据集。该数据集通过利用维基百科内部链接的锚文本及其上下文，自动收集了大量的事件共指提及，共有40,529条记录。数据集的创建过程涉及自动化的数据收集和手动验证，确保了数据的质量。WEC数据集主要包含参考性事件提及，适用于多文本信息匹配和集成等应用，旨在解决跨文档事件共指的挑战。

Wikipedia Event Coreference (WEC) dataset is a large-scale cross-document event coreference dataset developed by Bar-Ilan University in Israel. It automatically collects a substantial volume of event coreferential mentions by utilizing the anchor texts of internal Wikipedia links and their contextual surroundings, with a total of 40,529 records. The dataset construction process combines automated data collection and manual verification to guarantee data quality. The WEC dataset primarily consists of referential event mentions, making it applicable to scenarios such as multi-text information matching and integration, with the goal of addressing the challenges posed by cross-document event coreference.