Wikidata Knowledge Triples

该数据集包含了从Wikidata知识库精心挑选的13.2亿个知识三元组，专注于21个地理属性，旨在分析基于性别和年龄的职业偏见。数据集还包括了对人类实体敏感的属性，如年龄和性别，并经过筛选过程确保数据质量。该数据集规模宏大，包含1.32亿个三元组和9300万个实体，其任务旨在进行链接预测和偏见分析。

This dataset contains 132 million carefully curated knowledge triples selected from the Wikidata knowledge base, focusing on 21 geographic attributes, and is aimed at analyzing occupational biases based on gender and age. The dataset also includes sensitive attributes associated with human entities, such as age and gender, and has undergone a rigorous screening process to ensure data quality. Boasting a large scale, this dataset comprises 132 million triples and 93 million entities, with its core tasks including link prediction and bias analysis.