Annotated Corpus of Webtables

Annotated Corpus of Webtables是由阿尔伯塔大学创建的一个包含217,834个维基百科表格的数据集，这些表格被标注了28种关系。数据集通过远监督和精心设计的查询方法进行创建，旨在提高信息提取任务的准确性。该数据集的应用领域包括网络搜索和问答系统，旨在解决从结构化文档中提取实体和关系的问题。

The Annotated Corpus of Webtables is a dataset developed by the University of Alberta, which contains 217,834 Wikipedia tables annotated with 28 types of relations. This dataset was constructed via distant supervision and well-designed query methods, with the objective of enhancing the accuracy of information extraction tasks. Its application fields include web search and question answering systems, and it aims to address the challenge of extracting entities and relations from structured documents.

搜集汇总

背景与挑战

背景概述

该数据集是一个标注的网络表格语料库，包含28个Freebase关系的CSV和JSON文件，用于信息提取任务。数据集提供了丰富的表格和列号对应关系，适用于计算机与信息科学领域的研究。