CloneCognition: machine learning based code clone validation tool

Abstract

A code clone is a pair of code fragments, within or between software systems that are similar. Since code clones often negatively impact the maintainability of a software system, several code clone detection techniques and tools have been proposed and studied over the last decade. However, the clone detection tools are not always perfect and their clone detection reports often contain a number of false positives or irrelevant clones from specific project management or user perspective. To detect all possible similar source code patterns in general, the clone detection tools work on the syntax level while lacking user-specific preferences. This often means the clones must be manually inspected before analysis in order to remove those false positives from consideration. This manual clone validation effort is very time-consuming and often error-prone, in particular for large-scale clone detection. In this paper, we propose a machine learning approach for automating the validation process. First, a training dataset is built by taking code clones from several clone detection tools for different subject systems and then manually validating those clones. Second, several features are extracted from those clones to train the machine learning model by the proposed approach. The trained algorithm is then used to automatically validate clones without human inspection. Thus the proposed approach can be used to remove the false positive clones from the detection results, automatically evaluate the precision of any clone detectors for any given set of datasets, evaluate existing clone benchmark datasets, or even be used to build new clone benchmarks and datasets with minimum

Predictive Models in Software Engineering: Challenges and Opportunities

“…Since many clone detectors return code fragments that are not considered clones by users, it is necessary to perform manual validation of the reported possible clones. Mostaeen et al [176] present a tool called CloneCognition that can automate the laborious manual validation process by using Artiicial Neural Network (ANN). Their tool showed promising performance when compared with state-of-the-art clone validation techniques.…”

A Survey on Machine Learning Techniques for Source Code Analysis

“…Bandara and Wijayarathna [38] identified features such as the number of characters and words, identifier count, identifier character count, and underscore count using antlr tool. Some studies [221,223,224] utilized line similarity and token similarity. Yang et al [349] computed tf-idf along with other metrics such as position of clones in the file.…”

“…Similarly, Sheneamer and Kalita [288] compared the performance of Support Vector Machine, Linear Discriminant Analysis, Instance-Based Learner, Lazy K-means, Decision Tree, Naive Bayes, Multilayer Perceptron, and Logit Boost. dl-based models: dl models such as ann [223,224], dnn [99,365], and rnn with Reverse neural network [342] are also employed extensively. Bui et al [58] and Bui et al [57] combined neural networks for ml models training.…”

Predictive Models in Software Engineering: Challenges and Opportunities

“…On the other hand, creating clones in source code may introduce new defects or bugs, which can lead to extra maintenance effort to ensure consistency among cloned code snippets. Since software clones have a clear influence on software maintenance and evolution, many studies [85,90,118,139,145,146] proposed various approaches to detect and management code clones in source code.…”

“…They also applied Random Forest classifier to predict the life expectancy of newly-introduced clones. Mostaeen et al [85] present a tool called CloneCognition that can automate the laborious manual validation process by using Artificial Neural Network (ANN). Their tool showed promising performance when compared with state-of-the-art clone validation techniques.…”