Biomedical research is hampered by the triple challenges of disjointed data, unstructured data explosion, and lack of accessibility. In this talk, we demonstrate how combining new innovations in NLP and Graph Analytics can act as a potent remedy and accelerate biomedical research, by automatically building knowledge graphs from unstructured documents.
We introduce three techniques to fuse disjointed datasets, analyze them as a whole, surface hidden patterns, and answer concrete research questions.
We also present the biomedical knowledge graph we built in the process, the entities and relations in the database, and how they can be useful for various medical purposes such as clinical decision support and drug discovery.