Clustering dynamic class coupling data to measure class reusability pattern

Abstract

Identification of reusable components during the process of software development is an essential activity. Data mining techniques can be applied for identifying set of software components having dependence amongst each other. In this paper an attempt has been made to identify the group of classes having dependence amongst each other existing in the same repository. We explore document clustering technique based on tf-idf weighing to cluster classes from vast collection of class coupling data for particular java project/program. For this purpose firstly dynamic analysis of java application is done using UML diagrams to collect class import coupling data. Then in second step, this coupling data of each class is treated as a document and represented using VSM (using TF and IDF). Then finally in the third step basic K-mean clustering technique is applied to find clusters of classes. Further each cluster is ranked for its goodness. © 2011 Springer-Verlag.