摘要
在开发类社交平台上针对特定软件项目的讨论中往往蕴含着许多潜在的软件需求,针对GitHub中的Issues模块数据尝试对其进行安全性需求挖掘。通过抓取特定软件项目Issues中带有安全相关标签的讨论数据形成初选数据集,借助实体识别方法对句子中包含的连接实体进行识别,提取出用户需求,提出一种CreUS用户故事生成方法,将提取到的用户需求以用户故事的形式表示出来。实验结果表明,该方法生成的用户故事集可以有效支持项目开发,对于需求发现起到重要的辅助作用。
Discussions on specific software projects on development social media often contain many potential software requirements.Based on the Issues module data in GitHub,security requirements were mined.By fetching specific software project Issues related with security labels,the primary data set was formed,entity recognition technology was used to clarify the connecting entity in sentences,which identified and extracted the user demand,and a method,namely CreUS,was put forward to generate user story.The user requirements were presented in the form of user stories.Experimental results show that the user story set generated using the proposed method can effectively support the actual development of the project and play an important auxiliary role in requirement discovery.
作者
张晓妘
郑丽伟
ZHANG Xiao-yun;ZHENG Li-wei(Computer School,Beijing Information Science and Technology University,Beijing 100101,China)
出处
《计算机工程与设计》
北大核心
2022年第6期1628-1636,共9页
Computer Engineering and Design
基金
北京市自然科学基金项目(Z160002)。