Clustering-Based Support for Software Architecture Restructuring

Abstract

The maintenance of long-living software systems is an essential topic in today's software engineering practice and research. Software Architecture Restructuring is an important task to adjust these systems to current requirements and to keep them maintainable. Niels Streekmann introduces an approach to Software Architecture Restructuring that semi-automates this task by introducing graph clustering. The approach provides an iterative process that systematically incorporates human architectural knowledge for the improvement of the restructuring result. Thus, it supports the task of planning the transfer of an existing system to a target architecture and aims at reducing the required manual effort.