Hybrid Clustering Approach for software Module Clustering

  Kishore.C,  Dr.K.Ramani,  Anusha G

Maintenance is one of the main software creation activities in terms of allocated resources. As existing software ages, newly developed systems are built to improve upon existing systems. As software evolves, modularization structure of software degrades and at one point it becomes a challenging task to maintain the software future. Recently, clustering techniques have been used to help with the issues of software evolution and maintenance. It is well known fact that a good modularized software system is easy to understand and maintain. Software Module Clustering is an important task during the maintenance of the software whose main goal is to achieve good modularized software. In recent time, this problem has been converted into search based software engineering problem. All previous work on software module clustering used a single objective formulation of the problem. That is, twin objectives of high cohesion and low coupling have been combined into a single objective called modularization Quality. We introduced hybrid clustering approach which improves modular structure of the software system. We present the results and comparing the results obtained with this from existing single objective formulation on 7 real world model clustering problems. The results of this empirical study prove that hybrid clustering approach produces significantly better solutions than the existing single-objective approach.

Keywords- Search Based Software Engineering, Software Module Clustering, Modularization Quality.

Unique Identification Number - IJEDRCP1403035
Page Number(s) - 181-185
Pubished in - Volume 2 | Issue NCETSE Conference | March 2014
DOI (Digital Object Identifier) -   
Publisher - IJEDR (ISSN - 2321-9939)

  Kishore.C,  Dr.K.Ramani,  Anusha G,   "Hybrid Clustering Approach for software Module Clustering", International Journal of Engineering Development and Research (IJEDR), ISSN:2321-9939, Volume.2, Issue NCETSE Conference, pp.181-185, March 2014, Available at :http://www.ijedr.org/papers/IJEDRCP1403035.pdf