Developing a Novel and Effective Clone Detection Using Data Mining Technique

− A code clone is a code portion that is identical or similar to another in a source files. These code clones are called simple clones. And higher level similarities are called structural clones. Since several techniques have been proposed to detect this simple clone, still they occur in large software systems. This new algorithm detects the code clone for control structures such as for, while and do statements. First we extract the structural clone. Secondly we split the code clone from the original source code called source units and assign index values for each statement. Source units are then encoded in tuple form as one indicating the statements of source units and the other represents the index occurrence. Finally we calculate total number of occurrence for each index value. Keyswords− Code Clone, Cluster, Detection, Frequent, Itemset, Index.