We’ve recently had a conversation with a couple of teachers who said that they spend around 90% of their time analyzing students’ submissions to find out whether or not plagiarism has occurred.

Lengthy writing assignments and big class sizes make things even more complicated. This is what educators from across the globe are experiencing today.

The Problem of Last Minute Submissions ...

Part of the problem is that students tend to do assignments last minute due to poor writing skills, lack of motivation, or poor understanding of assignment instructions. This pushes them to look for easier solutions like unfair paraphrasing.

Extensive paraphrasing without due credit occurs less frequently because it takes more effort. On the contrary, the synonym swapping technique appears to be more common as it requires literally minutes to mask unoriginality.

So, struggling students would find text spinners online to substitute words with synonyms at random. As a result, they would get unfairly paraphrased texts that have more chances to bypass plagiarism checkers if they can catch exact matches only.

Not all the tools are equally effective, so educators are still capable of telling student-written sentences from those rephrased by the tools. Many times, these automatic spinners just replace words with ones that have nothing to do with the context. The submission content becomes simply impossible to read. But nonetheless, these tricks are sometimes enough to bypass plagiarism checkers.

Turnitin’s Algorithm Catches Synonym Swapping

The good news is that Turnitin’s Team has rolled out a great update to its similarity checking algorithm. It can recognize instances of synonym swapping and compare these illegally paraphrased sentences against Turnitin’s best-in-class content databases.

What does it mean to you, as an educator? You’ll receive Similarity Reports with highlighted text similarities including the sentences with swapped synonyms. This will refine the accuracy of the Similarity score and matches, resulting in a better experience for instructors reviewing submissions.

The algorithm has been implemented across Turnitin’s portfolio of products. This is an important step on Turnitin’s path toward continuing to identify and emerging trends in academic misconduct and how to make it easier for instructors to identify learning opportunities for students.

Please note:

If you encounter a suspected case of text spinning, please send it over to our team for further investigation. As technology continues to advance, having additional examples can help us to continue to improve our algorithm and capabilities to address evolving challenges.