CloneTracker: Advanced Dual-Layered Approach for Precise Detection of Clones in Java Applications

In the fast-paced world of software development, conventional code clone detection methods are becoming inadequate, frequently struggling to identify complex clones. This gap significantly endangers software quality and security, emphasizing the need for advanced clone detection approach. In response, our CloneTracker introduces an advanced, dual-layered solution to identify all types of clones. CloneTracker utilizes a dual-layer detection strategy based on function-level and chunk-level analysis for enhanced accuracy. Specifically, it prioritizes function-based signature detection, followed by chunk-based detection, requiring at least 50% similarity and three matching chunks from the same file for clone identification. The dual-level indexing and synergistic abstraction distinguishes our approach, surpassing traditional methods in efficiency and effectiveness. A comprehensive evaluation against well-known clone detection techniques across multiple reliable datasets revealed CloneTracker’s peerless precision average of 98%, recall ranging from 99% to 100%, accuracy ranging 97.5% and 99%, and efficiency of 1.14 to 1.29 seconds, with these results reflecting averages over function-based and file-based assessment. Its superior efficacy reaffirms the approach’s importance in modern software development, setting a new standard for clone detection and paving the way for future advancements. Beyond mere detection, CloneTracker also precisely identifies changes in cloned code, enhancing code quality and maintenance practices in the modern software development paradigm.