EVOLUTION STRATEGIES BASED AUTOMATED SOFTWARE CLUSTERING APROACH

Abstract

Maintenance is one of the key phases of software development life cycle, for long term effective use of any software. It can become very lengthy and costly for large software systems, especially when subsystem boundaries are not clearly defined. System evolution, lack of up to date documentation and high turnover rate of software professionals (leading to non availability of original designers of the software systems) can complicate the system structure many folds by making the subsystem boundaries ambiguous. Automated software module clustering helps software professionals to recover high-level structure of the system by decomposing the system into smaller manageable subsystems, containing interdependent modules. We treat software clustering as an optimization problem and propose a technique to get near optimal decompositions of relatively independent subsystems, containing interdependent modules. We propose the use of self adaptive Evolution Strategies to search a large solution space consisting of modules and their relationships. We compare our proposed approach with a widely used genetic algorithm based approach on a number of test systems. Our proposed approach shows considerable improvement in terms of quality and effectiveness and consistency of the solutions for all tests cases.