Copy detection for digital documents

Plagiarism is becoming a commonly concerned problem in recent years. With the advance of Internet, it is easier and easier to access the writings of other people. When someone uses the content of a writing inappropriately, it may cause the problem of plagiarism. Plagiarism may infringe the intellectual property rights, so it is a serious problem nowadays. In this paper, we propose a new method for detecting plagiarism. We use Word2vec to obtain the vector representations of the sentences involved in the documents. We also apply principal component analysis (PCA) to reduce the size of dimensionality. By using sentence-level representations, plagiarisms can be effectively detected. We will present experimental results and compare with other methods to show the plagiarism detection capability of our proposed method.