In software development, code clone can significantly reduce costs and speed up the development process. However, Indiscreet use of the clone not only lower the quality of the code but also adds extra cost and time to fix the bugs. Due to this problem...
In software development, code clone can significantly reduce costs and speed up the development process. However, Indiscreet use of the clone not only lower the quality of the code but also adds extra cost and time to fix the bugs. Due to this problem, many researchers have been trying to detect clones. However, the existing clone detection techniques are limited to detect clones that are fully identical or only with modified identifiers and they don't give information about clone type. This paper proposes a TBCNN(Tree-Based Convolution Neural Network)-based two-pass clone classification technique for detecting variety types of clones as well as their automatic classification. The experiment is performed using BigCloneBench, a well known and wildly adopted data set for clone detection. As a result, the proposed technique was able to classify four types of code clones with 78% recall and precision.