Improving The Detection of Plagiarism in Scientific Articles Using Machine Learning Approaches

One of the modern problems that occur in the current research and publication process is the duplication of the results of other people's research that is presented again by other parties. With the ease of the resources obtained, the more open the opportunity to bring up a problem called Plagiarism. This is attempted to be completed by the computer system with new approaches to detect and predict the existence of plagiarism in research automatically. In this article, approaches and methods for detecting plagiarism use machine learning techniques, where machine learning is empowered to become an algorithm as construction and evaluation in detecting plagiarism. Technically, this algorithm will compare and analyze the compatibility of words and sentences in documents with other document databases so that the analysis becomes an evaluation material, prediction, and determination that the document is plagiarism or not. The purpose of this study is to protect intellectual property and ideas, as well as the results to improve better performance and level of accuracy in detecting plagiarism.