GitHub Issue and PR Comments Bot Identification Dataset

本数据集由比利时蒙斯大学软件工程实验室创建，专注于GitHub上的Issue和PR评论，旨在通过手动分析识别机器人账户。数据集包含5000个独特的GitHub账户，其中527个被确认为机器人。该数据集用于开发和评估一个自动分类模型，以区分机器人和人类账户，主要通过分析账户的评论数量、模式和内容不平等性。此数据集的应用领域包括提高开源软件开发的安全性和效率，以及研究机器人对软件开发过程的影响。

This dataset was created by the Software Engineering Laboratory of the University of Mons, Belgium, with a focus on Issues and Pull Request (PR) comments on GitHub. Its objective is to identify bot accounts via manual analysis. The dataset comprises 5,000 unique GitHub accounts, 527 of which have been verified as bot accounts. This dataset is utilized for developing and evaluating an automated classification model to distinguish between bot and human accounts, mainly by analyzing the comment volume, patterns, and content inequality associated with these accounts. The application scope of this dataset includes improving the security and efficiency of open-source software development, as well as investigating the impact of bots on the software development process.