🤖 AI Summary
This study addresses the growing threat of “paper mills”—organized entities producing fraudulent scholarly articles—by proposing a quantifiable, reproducible detection framework leveraging Web of Science publication and citation data. Methodologically, it integrates conventional bibliometric indicators (e.g., h-index, citation counts) with a novel metric, the Integrity Index (I-index), which jointly models collaboration network topology, author productivity patterns, and citation anomalies to objectively identify low-quality, mass-produced papers. An automated MATLAB-based analytical tool implements the framework. Empirical evaluation demonstrates that the I-index achieves high discriminative power between paper mill–generated publications and legitimate scholarly output, yielding accurate detection across multiple confirmed cases. The framework establishes a scalable, quantitative paradigm for institutional and publisher-level academic integrity monitoring.
📝 Abstract
Based on the analysis of the data obtainable from the Web of Science publication and citation database, typical signs of possible papermilling behavior are described, quantified, and illustrated by examples. A MATLAB function is provided for the analysis of the outputs from the Web of Science. A new quantitative indicator -- integrity index, or I-index -- is proposed for using it along with standard bibliographic and scientometric indicators.