Clones clustering using K-means

Cloning is a process of reusing the existing code for development of fresh code or to modify an existing system. It involves using a known pattern or source code as aviation over which a new code designed with or without modifying the original source. Several approaches are being used for detection of clones. In our work we modified LSH base approach of Deckard to find clones. Deckard is a scalable and accurate clone detection tool which is LSH (Locality Sensitive Hashing) algorithm based. In this paper have we to replace the call to LSH with K-Means algorithm. LSH based proposed system will be used for of clones in Java, C, Php programs and will help in clone code optimization. K-Means algorithm for clustering uses set of observations to partition them into clusters.