Nonsensical computer-generated papers are still making their way into the scientific literature, Nature News reports.
In a new study appearing in the Journal of the Association for Information Science and Technology, the University of Toulouse's Guillaume Cabanac and the University Grenoble Alpes' Cyril Labbé developed and applied a tool aimed at uncovering papers generated using SCIgen, a MIT-designed software tool that strings together meaningless jargon. In 2014, Labbé, then at the Joseph Fourier University, uncovered more than 120 papers published by Springer and the Institute of Electrical and Electronic Engineers that had been cobbled together using SCIgen by analyzing certain words SCIgen uses.
He and Cabanac have now developed a new tool based on grammatical phrases to tease out SCIgen-generated papers. With it, they have uncovered 243 suspect papers within the Dimensions database, 46 of which had been retracted or removed.
Labbé tells Nature News that he suspects most of these papers were created to bulk up researchers' CVs and adds that it is a symptom of the "publish or perish" ways of academic life. "You shouldn't find these things in the literature," he tells it.