A clustering technique based on the specifications of software components

Software engineering is a major research area which can integrate other technical application areas into it. The important issue in software engineering is software reuse. It is always better to reuse rather than doing it from scratch. In order to attain the software reusability, an efficient design of software repository is needed. The searching algorithms and retrieval mechanisms will work effectively for a well defined software repository. In this paper, the data mining technique specifically clustering is applied for designing of software component repository. The components considered in our proposed algorithm are functions and the main focus is given to the specifications or behavior of the components. Here the specifications of software components are represented by using preconditions and postconditions. These preconditions and postconditions are converted to first order predicate logic and then the clustering is applied.