Text2RDF

Text2RDF是一个用于文本到RDF(资源描述框架)转换任务的专业数据集，包含从教育领域收集的1000段文本，通过人工标注生成了7228条高质量的微调数据。该数据集旨在支持自然语言处理模型学习如何将非结构化文本转换为结构化的RDF三元组表示。每条数据包含两个主要部分：1. input: 原始文本内容；2. output: 结构化RDF三元组表示，包含多个SPO(Subject-Predicate-Object)关系；每个SPO关系包含：• subject: 主体实体，包含名称、类型和属性• predicate: 谓词关系，描述主体和客体之间的关系• object: 客体实体，包含名称、类型和属性

Text2RDF is a specialized dataset for the text-to-RDF (Resource Description Framework) conversion task. It includes 1,000 text segments collected from the education domain, with 7,228 high-quality fine-tuning samples generated via manual annotation. This dataset aims to assist natural language processing (NLP) models in learning to convert unstructured text into structured RDF triple representations. Each data instance consists of two primary parts: 1. input: the original text content; 2. output: the structured RDF triple representation, which encompasses multiple SPO (Subject-Predicate-Object) relationships. Each SPO relationship contains: • subject: the subject entity, including its name, type and attributes • predicate: the predicate that describes the relational connection between the subject and the object • object: the object entity, including its name, type and attributes