IPRE

IPRE数据集由苏州大学计算机科学与技术学院创建，专注于人际关系提取，包含超过41,000条标注句子，涉及34种人际关系类型。数据集通过远监督方法生成，部分数据由人工标注，旨在提高人际关系识别的准确性，并支持知识图谱构建研究。IPRE数据集的应用领域包括信息提取和知识图谱构建，旨在解决现有知识库中人际关系不准确的问题。

The IPRE dataset was developed by the School of Computer Science and Technology, Soochow University, and focuses on interpersonal relation extraction. It contains over 41,000 annotated sentences covering 34 types of interpersonal relationships. The dataset is generated through the distant supervision method, with partial data manually annotated, aiming to enhance the accuracy of interpersonal relation recognition and support research on knowledge graph construction. Its application domains include information extraction and knowledge graph construction, and it is designed to address the issue of inaccurate interpersonal relations in existing knowledge bases.