Referency: Harmonizing Citations in Transdisciplinary Scholarly Literature

Elisa Bastianello, Alessandro Adamou, Nikos Minadakis

Abstract: The digital transformation of research articles that cross contemporary and traditional disciplines faces several common hurdles. The heterogeneous formats of bibliographical references, citations of works that predate DOIs, the inlining of references in notes and the use of co-referencing, diminish the effectiveness of modern-day citation extraction services. We combine structural and syntactic analysis of citations with the incorporation of knowledge bases to detect, extract and link citations and integrate them in structured texts. Its implementation, Referency, currently achieves what is unattainable even by contemporary machine learning-based solutions when dealing with nonstandard citations.