Because of the complexity of software development, some software developersmay plagiarize source code from other projects or open source software inorder to shorten development cycle. Many methods have been proposed to detectplagiarism among programs based on the program dependence graph, a graph representationof a program. However, to our best knowledge, existing works onlydetect similarity between programs without detecting copy direction among them. By employing extreme learning machine (ELM), we construct feature space fordescribing features of every two programs with possible plagiarism relationship. Such feature space could be large and time consuming, so we propose approachesto construct a small feature space by pruning isolated control statementsand removable statements from each program to accelerate both training and classificationtime. We also analyze the features of data dependencies between anyoriginal program and its copy program, and based on it we propose a feedbackframework to find a good feature space that can achieve both accuracy and efficiency. We conducted a thorough experimental study of this technique on real Cprograms collected from the Internet. The experimental results show the high accuracyand efficiency of our ELM-based approaches.
展开▼
