中国科学院机构知识库网格
Chinese Academy of Sciences Institutional Repositories Grid
network analysis of oss evolution: an empirical study on argouml project

文献类型:会议论文

作者Zhang Wen ; Yang Ye ; Wang Qing
出版日期2011
会议名称2011 12th International Workshop on Principles on Software Evolution and 7th ERCIM Workshop on Software Evolution, IWPSE-EVOL'11
会议日期September
会议地点Szeged, Hungary
页码71-80
中文摘要While complexity is an essential problem inherent in software system and its development, OSS (Open-Source Software) is not an exception and is not immune to this problem as well. The fast growth of OSS movement has impressed us with reduced cost but high quality software. To learn some lessons from successful OSS in handling the complexity, social network analysis is prevalent in analyzing both human-aspect and source-code-aspect interaction of OSS. This paper conducted an empirical study of an OSS project-ArgoUML. Unlike most previous studies regarding OSS email archives as a whole social network, our focus is on the quantitative analysis of a series of social networks produced in the process of OSS version evolution and module development. Through the empirical study, we have found that all the social network measures employed in this study are comparable to identify core developers of ArgoUML project. The frequency of co-occurrence of developers within the same topic is not a decisive factor to identify core developers. Developers within the same module communicate closely and frequently with each other. The more modules a developer developed, the more communication he (or she) will have with other developers. Although participants of developers' mailing lists are fluctuating in a large magnitude, the committers of the source code are kept stable in each version evolution. Moreover, the variation of committers of source code in version evolutions is almost unpredictable based on the variation of participants in developers' mailing lists. © 2011 ACM.
英文摘要While complexity is an essential problem inherent in software system and its development, OSS (Open-Source Software) is not an exception and is not immune to this problem as well. The fast growth of OSS movement has impressed us with reduced cost but high quality software. To learn some lessons from successful OSS in handling the complexity, social network analysis is prevalent in analyzing both human-aspect and source-code-aspect interaction of OSS. This paper conducted an empirical study of an OSS project-ArgoUML. Unlike most previous studies regarding OSS email archives as a whole social network, our focus is on the quantitative analysis of a series of social networks produced in the process of OSS version evolution and module development. Through the empirical study, we have found that all the social network measures employed in this study are comparable to identify core developers of ArgoUML project. The frequency of co-occurrence of developers within the same topic is not a decisive factor to identify core developers. Developers within the same module communicate closely and frequently with each other. The more modules a developer developed, the more communication he (or she) will have with other developers. Although participants of developers' mailing lists are fluctuating in a large magnitude, the committers of the source code are kept stable in each version evolution. Moreover, the variation of committers of source code in version evolutions is almost unpredictable based on the variation of participants in developers' mailing lists. © 2011 ACM.
收录类别EI
会议主办者Special Interest Group on Software Engineering (SIGSOFT); ERCIM
会议录IWPSE-EVOL'11 - Proceedings of the 12th International Workshop on Principles on Software Evolution
语种英语
ISBN号9781450308489
源URL[http://ir.iscas.ac.cn/handle/311060/16204]  
专题软件研究所_软件所图书馆_会议论文
推荐引用方式
GB/T 7714
Zhang Wen,Yang Ye,Wang Qing. network analysis of oss evolution: an empirical study on argouml project[C]. 见:2011 12th International Workshop on Principles on Software Evolution and 7th ERCIM Workshop on Software Evolution, IWPSE-EVOL'11. Szeged, Hungary. September.

入库方式: OAI收割

来源:软件研究所

浏览0
下载0
收藏0
其他版本

除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。