论文标题
一项关于读书我的内容与项目受欢迎程度之间相关性的实证研究
An Empirical Study On Correlation between Readme Content and Project Popularity
论文作者
论文摘要
GITHUB存储库中的Readme是信息的初步来源，因此可以帮助开发人员了解项目，重用或扩展。读书文件中存在不同类型的上下文和结构内容，我们将其称为内容和功能的类别和功能，并且可以确定对项目的理解程度。因此，内容的结构和上下文方面可能会影响项目的受欢迎程度。研究内容与项目受欢迎程度之间的相关性可以有助于关注可以提高流行度的各个方面，同时设计回能文件。但是，现有研究探讨了读书文件中内容和功能类型的类别，并且不探索它们对项目受欢迎程度的有用性。因此，我们提出了一项经验研究，以了解回复文件内容和项目受欢迎程度之间的相关性。我们对1950年公共GitHub项目的读书文件进行了研究，跨越了十种编程语言，并观察到大多数流行项目中的读数文件都是使用列表和图像井井有条的，并包含指向外部来源的链接。此外，观察到包含贡献指南的回复文件的存储库和参考文献与更高的受欢迎程度有关。
Readme in GitHub repositories serves as a preliminary source of information, and thus helps developers in understanding about the projects, for reuse or extension. Different types of contextual and structural content, which we refer to as categories of the content and features in the content respectively, are present in readme files, and could determine the extent of comprehension about project. Consequently, the structural and contextual aspects of the content could impact the project popularity. Studying the correlation between the content and project popularity could help in focusing on the aspects that could improve popularity, while designing the readme files. However, existing studies explore the categories of content and types of features in readme files, and do not explore their usefulness towards project popularity. Hence, we present an empirical study to understand correlation between readme file content and project popularity. We perform the study on 1950 readme files of public GitHub projects, spanning across ten programming languages, and observe that readme files in majority of the popular projects are well organised using lists and images, and comprise links to external sources. Also, repositories with readme files containing contribution guidelines and references were observed to be associated with higher popularity.
