TEKGEN-Wiki

<div style="text-align: center;"> <img src="textgen.jpg" alt="A person writing KGs on green board" width="250" height="250"> </div> **TEKGEN-wiki** is derived from the TEKGEN dataset released by [Google Research](https://github.com/google-research-datasets/KELM-corpus). TEKGEN is a corpus used for fine-tuning the T5-large model to improve Knowledge Graph (KG) generation ([NAACL 2021 Paper](https://aclanthology.org/2021.naacl-main.278.pdf)). This dataset provides the complete collection of original sentences from the TEKGEN dataset.<div style="text-align: center;"><img src="textgen.jpg" alt="一名在绿板上书写知识图谱（Knowledge Graph，KG）的人员" width="250" height="250"></div> **TEKGEN-wiki** 源自Google Research发布的TEKGEN数据集（代码仓库：https://github.com/google-research-datasets/KELM-corpus）。TEKGEN是用于微调T5-large模型以提升知识图谱（Knowledge Graph，KG）生成能力的语料库（相关论文：[NAACL 2021会议论文](https://aclanthology.org/2021.naacl-main.278.pdf)）。本数据集完整收录了TEKGEN数据集的全部原始语句。

搜集汇总

背景与挑战

背景概述

TEKGEN-Wiki数据集源自Google Research的TEKGEN语料库，专为微调T5-large模型以优化知识图谱生成而设计。它提供了TEKGEN数据集中所有原始句子的完整集合。