Comparison of Algorithms for the Recognition of ChatGPT Paraphrased Texts

The rapid development of artificial intelligence, especially chatbots, is leading to new forms of plagiarism that are difficult to detect using existing methods. Paraphrasing tools make this problem even more difficult and are incredible in minor languages with inadequate resources and tools. This study explores strategies that can help detect plagiarism generated by ChatGPT 4.0 and altered by paraphrasing tools. We propose two new datasets consisting of abstracts of doctoral theses in English and Serbian. Both datasets were subjected to ChatGPT paraphrasing, which allowed us to form two classes of texts: human-generated and AI-generated, i.e. AI-paraphrased. We then perform a comprehensive comparison of 19 widely used classification algorithms based on two feature sets, namely word unigrams and character multigrams. In addition, we compare these to the results of a commercially available pre-trained ChatGPT content detector, ZeroGPT. The results on the English corpus turn out to be very accurate, achieving an accuracy of 95% or more. In contrast, the results on the Serbian corpus were less accurate, achieving an accuracy of just over 85%. We attribute this difference to the lower ability of ChatGPT to parahprase in minor languages such as Serbian.