Mining Social Collaboration Patterns in Developer Social Networks.

Mohammed Abdelrahman Aljemabi,Zhongjie Wang,Mohammed A. Saleh
DOI: https://doi.org/10.1049/iet-sen.2019.0316
2020-01-01
IET Software
Abstract:Software development is extremely complex, requiring collaboration between teams and developers who collaborate on various tasks; these activities lead to the generation of an implicit developer social network (DSN). The authors' aim to understand the development process in terms of collaboration between developers. In this work, they conducted an empirical study on mining social collaboration patterns of DSNs for open source software projects based on an integrated approach involving the identification of global and local collaboration patterns among developers based on social network analysis. The bug tracking system-based DSN (BTS-DSN) is chosen as an example over the other DSNs since it incorporates larger collaboration activities and actors. The empirical results show that the DSNs, specifically BTS-DSN, exhibits three different coordination pattern levels (Plan, Aware, and Reflexive) based on their collaboration activities. The mean time to repair metric proves that the Reflexive level occupies the fastest bug fixing time, then the Plan level comes secondly, and lastly the Aware level. In addition, each level group shows different collaboration behaviours among developers; thus, this information can be useful as a resource for better understanding of developer collaboration and collaboration awareness.
What problem does this paper attempt to address?