Yet another paper on AI/NLP-powered knowledge graph creation
Elsevier’s “Artificial Intelligence in the Life Sciences” is to publish another text from Fraunhofer SCAI with contributions from Kairntech about “A natural language processing system for the efficient updating of highly curated pathophysiology mechanism knowledge graphs”. Lead author Negin Babaiha and her colleagues outline the architecture of a system implemented at SCAI that uses evidence derived from the automatic processing of large amounts of scientific content to enrich and update structure knowledge repositories such as knowledge graphs.
A key ingredient in the described setup is the Kairntech NLP platform as coauthors Stefan Geissler and Bruno Freudensprung from Kairntech explain.
Fraunhofer SCAI and Kairntech have been partners on this issue for the last two years with Kairntech software being available on SCAI’s large parallel computing cluster, precisely for the purpose of generating large, indication-wide knowledge graphs from scientific content. Just recently, a large EU-funded research project from SCAI, Kairntech and a larger group of other research institutions has been accepted for funding.