A Method for Identifying and Recommending Reconstructed Clones

Abstract

Reconstruction of existing clone code is limited to a single version of static analysis and ignores the evolution of cloned code, resulting in a lack of effective methods for cloning code refactoring decisions. Therefore, this paper proposes a clone code that needs to be reconstructed and tracked from the perspective of software evolution, and recommends cloned code that needs to be reconstructed. Firstly, the evolution history information closely related to the cloned code is extracted from the clone detection, clone mapping, clone family and software maintenance log management system. Secondly, the clone code that needs to be reconstructed is identified, and the cloned code of the trace is identified, and then extracted and weighted. Construct related static and evolution features and build a feature sample database. Finally, a variety of machine learning methods are used to compare and select the best classifiers to recommend refactoring clones. This article conducts experiments on nearly 170 versions of 7 software, and recommends that the accuracy of refactoring cloned code reaches over 90%. Provide more accurate and reasonable code refactoring advice for software development and maintenance personnel.