The measurable increase in fake science and fraudulent publication practices – not least due to the availability of generative language models – is a threat to scientific integrity and jeopardizes trust in scientific communication. With the FACTS project (Fraud Analysis and Counterfeit Testing in Scholarly Publishing), the KOBV is attempting to develop countermeasures to this trend.
FACTS focuses on the development of graph-based models, generative AI models, and hybrid methods for analyzing large bibliometric datasets. The aim is to incorporate these methods and models to identify the mechanisms and practices behind predatory journals, paper mills, or citation cartels and to develop evidence-based decision-making aids with concrete options for action for publication consulting in Berlin libraries.
