After clicking the "Check it!" button, the text is passed to a word2vec model that transforms the text into a vector, which is then fed to a machine learning classifier. The classifier was trained on 8,024 document vectors of Wikipedia articles and their spun counterparts. The classifier can identify words that occur in an unusual context, which indicates the possible use of an online paraphrasing tool. For more details, please refer to the paper below.
If you find this tool useful, you may cite our paper:
Tomáš Foltýnek, Terry Ruas, Philipp Scharpf, Norman Meuschke, Moritz Schubotz, William Grosky and Bela Gipp. 2020. Detection of Machine-generated Plagiarism. In Proceedings of the iConference, March 23–26, 2020, Borås, Sweden.
Tomáš Foltýnek - tomas.foltynek@mendelu.cz
Terry Ruas - truas@umich.edu