An Effective Semantic Code Clone Detection Framework Using Pairwise Feature Fusion

Code clones. In this work, we propose a novel detection framework using machine learning for automated of all four type The features extracted from pair code blocks are combined possible clone with respect to reference block. We use AST and PDG both prepare labelled training samples after fusing the two feature vectors three different alternatives. six state-of-the-art classification models including Deep Convolutional Neural Network assess prediction performance our scheme. To access effectiveness seven datasets compare its five detectors. also large number algorithms detection. Comparing techniques, ANN non-ANN, such features, establishing that gives competitive results deep as well boosted tree algorithms, find like XGBoost quite in Experimental demonstrate approach outperforms existing methods terms accuracy.