Data for: Finding key classes in object-oriented software systems by techniques based on static analysis

This dataset accompanies the article: Ioana Sora, Ciprian-Bogdan Chirila, "Finding key classes in object-oriented software systems by techniques based on static analysis". In the article, we define different class attributes and investigate which attributes can be used best as a measure of class importance. We experiment on 15 open source software systems. This dataset contains: A list of the systems used as case studies is given as a text file. For every system, we specify the number of the used version and URLs to code and documentation. For every software system, the dataset further contains: - the reference solution as a set of known key classes, extracted from design documentation; - for all classes of the system, the values of all class attributes computed by us with the methods described in the article; - the positions on which the known key classes are ranked by the 3 approaches proposed in the article. Also supplementary data behind the plots is given in the form of 3 tables containing the raw data behind the boxplots in Figures 3, 4 and 5 of the article.本数据集配套Ioana Sora与Ciprian-Bogdan Chirila合著的论文《基于静态分析技术的面向对象软件系统关键类识别》（原英文标题："Finding key classes in object-oriented software systems by techniques based on static analysis"）。 在该论文中，我们定义了多种类属性，并探究哪些属性最适合作为类重要性的衡量指标。我们在15个开源软件系统上开展了实验。 本数据集包含以下内容： 1. 一份以文本文件形式存储的研究案例系统列表。针对每个系统，我们标注了其所用版本号，以及代码与文档的访问URL。 2. 针对每个软件系统，数据集还包含： - 参考解决方案：即从设计文档中提取的已知关键类集合； - 针对系统内所有类，通过论文所述方法计算得到的全部类属性取值； - 本文提出的3种方法对已知关键类的排序位次。 此外，数据集还以3张表格的形式提供了论文中图3、图4与图5箱线图对应的原始补充数据。