Requirement Relation Extraction Dataset
收藏资源简介:
本数据集名为Requirement Relation Extraction Dataset,由慕尼黑工业大学等机构创建,包含2093条来自需求工程领域的预处理句子。数据集大小适中,涵盖了从政府机构软件到视频游戏等多种主题的需求文档。创建过程中,研究人员手动提取并预处理了这些句子,确保其语法和拼写正确,同时去除了不必要的信息。该数据集主要用于关系抽取研究,旨在通过半自动标注框架提高标注效率和一致性,解决传统人工标注中的偏差和不一致问题。
This dataset is named Requirement Relation Extraction Dataset, and it was developed by institutions including the Technical University of Munich. It contains 2093 preprocessed sentences from the domain of Requirements Engineering. With a moderate scale, this dataset covers requirement documents across a wide range of topics, spanning from government agency software to video games. During the dataset construction process, researchers manually extracted and preprocessed these sentences to guarantee correct grammar and spelling, while eliminating redundant information. This dataset is primarily intended for relation extraction research, with the goal of enhancing annotation efficiency and consistency through a semi-automated annotation framework, thereby addressing the biases and inconsistencies associated with traditional manual annotation.




