In a recent article for Nature, Dalmeet Singh Chawla discusses a new approach to detecting fake research papers. While some approaches look for “tortured phrases” or identical manuscripts, this new approach analyses the relationships between authors to identify unusual patterns. Simon Porter, who developed the technology, noted that younger researchers typically publish with their supervisor before beginning to publish with other people; Work produced by a paper mill often comes from networks of early-career researchers with a sudden spike of activity or groups of authors who have no history collaborating and may be from different disciplines or locations. The papers that were flagged by this new software saw significant overlap with papers that were flagged by other technology.
