Natural Language Processing to Assess and Improve Citation Integrity in Biomedical Publications

This project will assist researchers and journals in evaluating citation behavior in biomedical publications. While citations play a fundamental role in the diffusion of scientific knowledge and assessment of research on a topic, they are often inaccurate (e.g., citation of nonexistent findings, inappropriate interpretation). This inaccuracy undermines the integrity of scientific literature and distorts the perception of available scientific evidence. Kilicoglu and Schneider will develop and validate resources and natural language processing/artificial intelligence models to aid stakeholders in assessing biomedical publications for citation accuracy and integrity. Stakeholders who will benefit from these new models and tools include authors, journals and peer reviewers, research administrators, funders, and policymakers. Read more.