A Study of Software Clone Detection Techniques for Better Software Maintenance and Reliability

Major problem in the development of software development is the presence of duplicate code that has a great impact on the overall affect the maintainability of the software. Clone detection technique applied with the core objective to identify the software codes which are identical. Various approached had been proposed in past by various researchers based which are based on text based, and token-based techniques. However, the proposed approaches were not considered very reliable methods due to the inability to find out syntactic differences between programs. Highly effective way to identify syntactic difference is through usage of abstract syntax tree. There are numerous ways to find similarity between two programs. In the proposed work, proposed software clone detection technique software code would be represented in the form of metrics affecting maintenance and reliability of opensource software. Afterwards, features extraction would be done in the form of flexible vectors of different forms to detect different types of clones. Proposed technique would be based upon adaptive prefix filtering on sets of vectors to detect similarity among the vectors. Similarity index detected among the vectors would be used to define given codes as code clones.