NED-SERIES
How to distill knowledge from biomedical text combining pre-trained language models with graph machine learning
This article synthesizes a paper accepted for the IEEE Application of Information and Communication Technologies (AICT2024) conference. In addition to the undersigned, Felice Paolo Colliani (first author), Giovanni Garifo, Antonio Vetrò, and Juan Carlos De Martin are the co-authors of this paper.
The biomedical domain has seen a steadily increasing publication rate over the years due to the growth of scientific research, advances in technology, and the global emphasis on healthcare and medical research.
The application of Natural Language Processing (NLP) techniques in the biomedical domain represents a shift in the analysis and interpretation of the vast corpus of biomedical knowledge, enhancing our ability to derive meaningful insights from textual data.
Named Entity Disambiguation (NED) is a critical NLP task that involves resolving ambiguities in entity mentions by linking them to the correct entries in a knowledge base. To understand the importance and complexity of such a task, consider the following example:
Zika belongs to the Flaviviridae family and it is spread by Aedes mosquitoes.
Individuals affected by Zika infection often…