Using automated analysis of the bibliography to detect potential research integrity issues

Abstract

Efforts to assess research integrity in scholarly publications tend to focus on the front matter (e.g., fake authors) and the article body (e.g., data falsification, image manipulation, plagiarism). Automated analysis of the bibliography may surface fingerprints pointing to research integrity issues. Tools exist that can be used at all workflow stages, from manuscript submission through production, to detect citations of retracted articles and articles published by predatory publishers. Validation of bibliographic references against trusted databases can be used to help detect fake or chimeric references generated by artificial intelligence applications.