On the development of a plagiarism detection model based on citation analysis using a bibliographic database

Abstract

A relatively new method for the detection of text plagiarism is proposed based on a search for original sources with an identical or similar list of references. First of all, this is applied to the most difficult to detect forms of translated plagiarism and the plagiarism of ideas. This method successfully proved itself in test studies of groups of foreign authors and is a continuation of our studies. The peculiarity of the approach that is proposed by the authors is the use of multidisciplinary bibliographic databases rather than full-text databases, which are often unavailable due to the high price of subscriptions under Russian conditions. The advantage of their use is the access to the maximum possible number of article references, which significantly extends the base for the search for originals while analyzing suspicious scientific publications. The step-by-step algorithm for addressing a query to the Web of Science and Scopus databases and analysis of the obtained data can be recommended for implementation into systems of plagiarism detection as an additional component.