EventEA
收藏资源简介:
EventEA是一个针对事件中心知识图谱的实体对齐数据集,由南京大学软件新技术国家重点实验室创建。该数据集包含20000个事件实体对齐对,主要从EventKG、DBpedia和Wikidata中提取,具有较少的同构结构和更多的异构属性值。EventEA旨在评估嵌入式实体对齐方法在处理困难实体对齐设置时的鲁棒性,特别适用于评估结构化方法和属性增强方法。数据集的应用领域包括知识图谱的跨语言和跨领域实体对齐,旨在解决现有数据集偏差和评估挑战。
EventEA is an entity alignment dataset for event-centric knowledge graphs, created by the State Key Laboratory for Novel Software Technology at Nanjing University. This dataset contains 20,000 event entity alignment pairs, mainly extracted from EventKG, DBpedia and Wikidata, and features fewer isomorphic structures and more heterogeneous attribute values. EventEA aims to evaluate the robustness of embedding-based entity alignment methods when handling challenging entity alignment scenarios, and is particularly suitable for assessing structured and attribute-enhanced alignment methods. The application scope of this dataset covers cross-lingual and cross-domain entity alignment on knowledge graphs, and it is designed to address existing dataset biases and evaluation challenges.




