Community structure of a complex software-system in evolution

Due to the discoveries of complex networks in the real-world systems, recent researches in software engineering have also applied the same principles in analysis of software-systems. Moreover, network analysis enabled the determination of different statistical metrics proving small-world phenomena, power-law and scale-free properties also for the software-systems. But also, statistical network analysis of software-systems reveals interesting community structure, which has not been already enough explored for software-systems. Since many real-world networks exhibit natural division into communities, the same behavior has been explored for the software network. In our research, we have analyzed large and complex evolving open-source software-system. For determining community structure, a statistical metric called modularity was used. The presence of structural modules proves a deeper organizing structure of the software-system, which can be applied for the prediction of many software characteristics.