CitePlag is a prototype of a hybrid Plagiarism Prevention and Detection System cooperatively develoepd by the Information Science Group at the University of Konstanz and Prof. Joeran Beel at the Trinity College Dublin.
CitePlag implements the Citation-based Plagiarism Detection (CbPD) approach, which was initially introduced in the doctoral thesis of Bela Gipp. Details on the algorithms implemented in CitePlag can be found here. The current prototype was developed in cooperation with students from the HTW, Berlin
Compared to existing approaches for plagiarism detection, the CitePlag prototype does not consider textual similarity alone, but uses citation patterns within academic documents as a unique, language-independent fingerprint to identify semantic similarity. This feature for the first time enables automated detection of strongly disguised plagiarism forms, including paraphrases, translated plagiarism, and even idea plagiarism.
The suitability of the CbPD approach in detecting disguised plagiarism was first demonstrated on the plagiarized thesis of former German defense minister Karl- Theodor zu Guttenberg [PDF]. While conventional detection approaches could not identify a single instance of translated plagiarism in the thesis, the CbPD approach detected 13 of the 16 translated plagiarisms. The effectiveness of the method was further demonstrated when applied to the works of multiple authors and various plagiarism styles in the VroniPlag collection. Evaluations of real-world plagiarism showed that plagiarists commonly disguise academic misconduct by paraphrasing copied text, but often do not substitute or rearrange the citations copied from the source document.
Most recently, the practicability of the CbPD approach was demonstrated by analyzing 185,000 publications in the comprehensive bioscience full-text database PubMed Central. The CbPD algorithms allowed the identification of several plagiarism cases that were non-machine-detectable using today’s prevalent methods. [PDF]. As a result, several publications were retracted, including a fraudulent medical study.
While the CbPD approach offers unique benefits, it should be seen as a supplement not a replacement to existing software-based plagiarism detection methods, since text-based and citation-based plagiarism detection approaches complement each other. The CitePlag prototype represents a visualization of concepts and algorithms developed by the Information Science Group around the idea of Citation-based Plagiarism Detection (CbPD).
For questions or comments feel free to contact us.