YAGO5M

该数据集是从YAGO 4中筛选出的一部分，包含了1292万条三元组、482万个实体和124个谓词。为了提高与用户查询的相关性，数据集经过处理，去除了查询中未出现的实体和谓词。其规模达到了1292万条三元组，任务旨在完成知识图谱的补全工作。

This dataset is a filtered subset from YAGO 4, containing 12.92 million triples, 4.82 million entities, and 124 predicates. To enhance relevance to user queries, the dataset was processed to eliminate entities and predicates that do not appear in any user queries. The core task of this dataset is knowledge graph completion.