ETMQ: A novel objective function — Automatic clustering of software intensive systems

Software intensive systems are increasingly becoming complex and difficult to maintain the system structure and its understandability due to extremely dynamic requirements. Automatic software clustering is an important research area in the software reverse engineering domain, which addresses this issue by decomposing the system into multiple subsystems (clusters) with related interdependent components for better understandability and manageability. Researchers have applied meta-heuristics to obtain near optimal solution considering this problem as a graph partitioning problem which is a NP-hard problem. In this work a novel objective function called “Enhanced Turbo MQ (ETMQ)” is proposed and GA is applied and its effectiveness is ascertained for medium size software systems.