一种基于BP神经网络的代码相似性检测方法

Code Similarity Detection Approach Based on Back-propagation Neural Network

如何有效地检测程序设计课程作业中的抄袭现象是一个重要的问题。传统的抄袭检测方法主要利用代码的属性或结构信息来度量代码之间的相似性。给出了一种基于误差反向传播(BP算法)多层前向神经网络的代码抄袭检测方法。提取程序之间的7种比较特征作为神经网络的输入,经过网络计算后得出程序的相似值,并将该值与抄袭决策阈值相比较以判定存在抄袭现象的程序集。实验结果表明,本方法具有很好的检测效果。

It is very important to find plagiarized programs in the field of computer science education. Traditional methods for program similarity use attribute counting or structure information to detect plagiarism. This paper presented a program similarity detection approach based on back propagation （BP algorithm） multi-layer feed-forward neural net works. We extracted seven compared features of the code as the input of the neural network, and obtained the program similarity through the network calculation. Comparing the result with the threshold value, we can find all groups of simlar programs. Experimental results show that our method is effective.