Resumen |
According to statistics, over the past year, the quality of education has fallen due to the pandemic, and the percentage of plagiarism in the work of students has increased. Modern plagiarism detection systems work well with external plagiarism, they allow to weed out works and answers that completely copy someone else’s published ideas. Using natural language processing methods, the proposed algorithm allows not only detecting plagiarism, but also correctly classifies students’ responses by the amount of plagiarism. This research paper implements a two-step plagiarism detection algorithm. In the experiment, the text was converted into a vector form by the GloVe method, and then segmented by K-means and the result was obtained by the FP-Growth unsupervised learning algorithm. © 2021, Springer Nature Switzerland AG. |