one tagger, many uses - illustrating the power of ontologies in named entity recognition
TRANSCRIPT
Lars Juhl Jensen
One tagger, many usesIllustrating the power of ontologies in named
entity recognition
C++ tagger
flexible matching
>1000 abstracts / second
inherently thread-safe
comprehensive dictionary
expansion rules
curated blacklist
80–90% precision
70–80% recall
molecular entities
genes/proteins
chemicals
biomedical literature
assess studiedness
association networks
co-occurrence
data integration
Szklarczyk et al., Nucleic Acids Research, 2015string-db.org
Kuhn et al., Nucleic Acids Research, 2016stitch-db.org
localization & expression
GO cellular component
Brenda Tissue Ontology
Binder et al., Database, 2014compartments.jensenlab.org
tissues.jensenlab.org Santos et al., PeerJ, 2015
diseases & side effects
Disease Ontology
diseases.jensenlab.org Frankild et al., Methods, 2015
MedDRA
FDA product labels
Kuhn et al., Nucleic Acids Research, 2016sideeffects.embl.de
Danish dictionaries
electronic medical records
post-market monitoring
organisms & habitats
NCBI Taxonomy
BioNLP shared task
Environment Ontology
EOL descriptions
Pafilis et al., Bioinformatics, 2016environments.hcmr.gr / eol.org
interactive annotation
Pafilis et al., Proceedings of BioCreative V, 2015extract.hcmr.gr
Pafilis et al., Proceedings of BioCreative V, 2015extract.hcmr.gr
BioCreative V
AcknowledgmentsSune Pletscher-
FrankildEvangelos PafilisDamian Szklarczyk
Christian von MeringMichael KuhnPeer Bork
Robert ErikssonPeter Bjødstrup
JensenSøren Brunak