<s>
Biomedical	B-General_Concept
text	I-General_Concept
mining	I-General_Concept
(	O
including	O
biomedical	O
natural	B-Language
language	I-Language
processing	I-Language
or	O
BioNLP	B-General_Concept
)	O
refers	O
to	O
the	O
methods	O
and	O
study	O
of	O
how	O
text	B-Algorithm
mining	I-Algorithm
may	O
be	O
applied	O
to	O
texts	O
and	O
literature	O
of	O
the	O
biomedical	O
and	O
molecular	O
biology	O
domains	O
.	O
</s>
<s>
As	O
a	O
field	O
of	O
research	O
,	O
biomedical	B-General_Concept
text	I-General_Concept
mining	I-General_Concept
incorporates	O
ideas	O
from	O
natural	B-Language
language	I-Language
processing	I-Language
,	O
bioinformatics	O
,	O
medical	O
informatics	O
and	O
computational	O
linguistics	O
.	O
</s>
<s>
The	O
strategies	O
developed	O
through	O
studies	O
in	O
this	O
field	O
are	O
frequently	O
applied	O
to	O
the	O
biomedical	O
and	O
molecular	O
biology	O
literature	O
available	O
through	O
services	O
such	O
as	O
PubMed	B-Library
.	O
</s>
<s>
Applying	O
text	B-Algorithm
mining	I-Algorithm
approaches	O
to	O
biomedical	O
text	O
requires	O
specific	O
considerations	O
common	O
to	O
the	O
domain	O
.	O
</s>
<s>
Large	O
annotated	O
corpora	O
used	O
in	O
the	O
development	O
and	O
training	O
of	O
general	O
purpose	O
text	B-Algorithm
mining	I-Algorithm
methods	O
(	O
e.g.	O
,	O
sets	O
of	O
movie	O
dialogue	O
,	O
product	O
reviews	O
,	O
or	O
Wikipedia	O
article	O
text	O
)	O
are	O
not	O
specific	O
for	O
biomedical	O
language	O
.	O
</s>
<s>
Resources	O
designed	O
to	O
aid	O
in	O
building	O
new	O
biomedical	B-General_Concept
text	I-General_Concept
mining	I-General_Concept
methods	O
have	O
been	O
developed	O
through	O
the	O
Informatics	O
for	O
Integrating	O
Biology	O
and	O
the	O
Bedside	O
(	O
i2b2	O
)	O
challenges	O
and	O
biomedical	O
informatics	O
researchers	O
.	O
</s>
<s>
Text	B-Algorithm
mining	I-Algorithm
researchers	O
frequently	O
combine	O
these	O
corpora	O
with	O
the	O
controlled	O
vocabularies	O
and	O
ontologies	O
available	O
through	O
the	O
National	O
Library	O
of	O
Medicine	O
's	O
Unified	O
Medical	O
Language	O
System	O
(	O
UMLS	O
)	O
and	O
Medical	O
Subject	O
Headings	O
(	O
MeSH	O
)	O
.	O
</s>
<s>
Like	O
other	O
text	O
documents	O
,	O
biomedical	O
documents	O
contain	O
unstructured	B-Application
data	I-Application
.	O
</s>
<s>
Both	O
unstructured	B-Application
text	O
and	O
semi-structured	O
document	O
elements	O
,	O
such	O
as	O
tables	O
,	O
may	O
contain	O
important	O
information	O
that	O
should	O
be	O
text	O
mined	O
.	O
</s>
<s>
Without	O
specific	O
adaptations	O
,	O
text	B-Algorithm
mining	I-Algorithm
approaches	O
designed	O
to	O
identify	O
claims	O
within	O
text	O
may	O
mis-characterize	O
these	O
"	O
hedged	O
"	O
statements	O
as	O
facts	O
.	O
</s>
<s>
Biomedical	B-General_Concept
text	I-General_Concept
mining	I-General_Concept
applications	O
developed	O
for	O
clinical	O
use	O
should	O
ideally	O
reflect	O
the	O
needs	O
and	O
demands	O
of	O
clinicians	O
.	O
</s>
<s>
This	O
is	O
a	O
concern	O
in	O
environments	O
where	O
clinical	B-Application
decision	I-Application
support	I-Application
is	O
expected	O
to	O
be	O
informative	O
and	O
accurate	O
.	O
</s>
<s>
New	O
text	B-Algorithm
mining	I-Algorithm
systems	O
must	O
work	O
with	O
existing	O
standards	O
,	O
electronic	B-Application
medical	I-Application
records	I-Application
,	O
and	O
databases	O
.	O
</s>
<s>
Text	B-Algorithm
mining	I-Algorithm
systems	O
operating	O
with	O
private	O
medical	O
data	O
must	O
respect	O
its	O
security	O
and	O
ensure	O
it	O
is	O
rendered	O
anonymous	O
where	O
appropriate	O
.	O
</s>
<s>
Developments	O
in	O
biomedical	B-General_Concept
text	I-General_Concept
mining	I-General_Concept
have	O
incorporated	O
identification	O
of	O
biological	O
entities	O
with	O
named	B-General_Concept
entity	I-General_Concept
recognition	I-General_Concept
,	O
or	O
NER	O
.	O
</s>
<s>
Most	O
entity	O
recognition	O
methods	O
are	O
supported	O
by	O
pre-defined	O
linguistic	O
features	O
or	O
vocabularies	O
,	O
though	O
methods	O
incorporating	O
deep	B-Algorithm
learning	I-Algorithm
and	O
word	B-General_Concept
embeddings	I-General_Concept
have	O
also	O
been	O
successful	O
at	O
biomedical	O
NER	O
.	O
</s>
<s>
Biomedical	O
documents	O
may	O
be	O
classified	B-Algorithm
or	O
clustered	O
based	O
on	O
their	O
contents	O
and	O
topics	O
.	O
</s>
<s>
These	O
two	O
tasks	O
are	O
representative	O
of	O
supervised	B-General_Concept
and	O
unsupervised	B-General_Concept
methods	O
,	O
respectively	O
,	O
yet	O
the	O
goal	O
of	O
both	O
is	O
to	O
produce	O
subsets	O
of	O
documents	O
based	O
on	O
their	O
distinguishing	O
features	O
.	O
</s>
<s>
Methods	O
for	O
biomedical	O
document	O
clustering	O
have	O
relied	O
upon	O
k-means	B-Algorithm
clustering	I-Algorithm
.	O
</s>
<s>
Text	B-Algorithm
mining	I-Algorithm
methods	O
may	O
perform	O
relation	O
discovery	O
to	O
identify	O
these	O
connections	O
,	O
often	O
in	O
concert	O
with	O
named	B-General_Concept
entity	I-General_Concept
recognition	I-General_Concept
.	O
</s>
<s>
In	O
practice	O
,	O
this	O
process	O
involves	O
both	O
isolating	O
phrases	O
and	O
sentences	O
denoting	O
the	O
core	O
arguments	O
made	O
by	O
the	O
authors	O
of	O
a	O
document	O
(	O
a	O
process	O
known	O
as	O
argument	B-Algorithm
mining	I-Algorithm
,	O
employing	O
tools	O
used	O
in	O
fields	O
such	O
as	O
political	O
science	O
)	O
and	O
comparing	O
claims	O
to	O
find	O
potential	O
contradictions	O
between	O
them	O
.	O
</s>
<s>
Information	B-General_Concept
extraction	I-General_Concept
,	I-General_Concept
or	I-General_Concept
IE	I-General_Concept
,	O
is	O
the	O
process	O
of	O
automatically	O
identifying	O
structured	O
information	O
from	O
unstructured	B-Application
or	O
partially	O
structured	O
text	O
.	O
</s>
<s>
IE	O
processes	O
can	O
involve	O
several	O
or	O
all	O
of	O
the	O
above	O
activities	O
,	O
including	O
named	B-General_Concept
entity	I-General_Concept
recognition	I-General_Concept
,	O
relationship	O
discovery	O
,	O
and	O
document	B-Algorithm
classification	I-Algorithm
,	O
with	O
the	O
overall	O
goal	O
of	O
translating	O
text	O
to	O
a	O
more	O
structured	O
form	O
,	O
such	O
as	O
the	O
contents	O
of	O
a	O
template	O
or	O
knowledge	O
base	O
.	O
</s>
<s>
Biomedical	B-General_Concept
text	I-General_Concept
mining	I-General_Concept
supports	O
applications	O
for	O
identifying	O
documents	O
and	O
concepts	O
matching	O
search	O
queries	O
.	O
</s>
<s>
Search	O
engines	O
such	O
as	O
PubMed	B-Library
search	O
allow	O
users	O
to	O
query	O
literature	O
databases	O
with	O
words	O
or	O
phrases	O
present	O
in	O
document	O
contents	O
,	O
metadata	O
,	O
or	O
indices	B-Data_Structure
such	O
as	O
MeSH	O
.	O
</s>
<s>
Similar	O
approaches	O
may	O
be	O
used	O
for	O
medical	B-Application
literature	I-Application
retrieval	I-Application
.	O
</s>
<s>
For	O
more	O
fine-grained	O
results	O
,	O
some	O
applications	O
permit	O
users	O
to	O
search	O
with	O
natural	B-General_Concept
language	I-General_Concept
queries	I-General_Concept
and	O
identify	O
specific	O
biomedical	O
relationships	O
.	O
</s>
<s>
On	O
16	O
March	O
2020	O
,	O
the	O
National	O
Library	O
of	O
Medicine	O
and	O
others	O
launched	O
the	O
COVID-19	O
Open	O
Research	O
Dataset	O
(	O
CORD-19	O
)	O
to	O
enable	O
text	B-Algorithm
mining	I-Algorithm
of	O
the	O
current	O
literature	O
on	O
the	O
novel	O
virus	O
.	O
</s>
<s>
Other	O
participants	O
include	O
Google	B-Application
,	O
Microsoft	O
Research	O
,	O
the	O
Center	O
for	O
Security	O
and	O
Emerging	O
Technology	O
,	O
and	O
the	O
Chan	O
Zuckerberg	O
Initiative	O
.	O
</s>
<s>
YesBioInferPyysalo	O
et	O
al.1	O
,	O
100	O
sentences	O
from	O
biomedical	O
research	O
abstracts	O
annotated	O
for	O
relationships	O
,	O
named	B-General_Concept
entities	I-General_Concept
,	O
and	O
syntactic	O
dependencies.NoBioScopeVincze	O
et	O
al.1	O
,	O
954	O
clinical	O
reports	O
,	O
9	O
papers	O
,	O
and	O
1,273	O
abstracts	O
annotated	O
for	O
linguistic	O
scope	O
and	O
terms	O
denoting	O
negation	O
or	O
uncertainty.YesBioText	O
Recognizing	O
Abbreviation	O
DefinitionsBioText	O
Project	O
1,000	O
abstracts	O
on	O
the	O
subject	O
of	O
"	O
yeast	O
"	O
,	O
annotated	O
for	O
abbreviations	O
and	O
their	O
meanings	O
.	O
</s>
<s>
Includes	O
annotations	O
of	O
clauses.YesIEPADing	O
et	O
al.486	O
sentences	O
from	O
biomedical	O
research	O
abstracts	O
annotated	O
for	O
pairs	O
of	O
co-occurring	O
chemicals	O
,	O
including	O
proteins.NoJNLPBA	O
corpusKim	O
et	O
al.An	O
extended	O
version	O
of	O
version	O
3	O
of	O
the	O
GENIA	O
corpus	O
for	O
NER	O
tasks.NoLearning	O
Language	O
in	O
Logic	O
(	O
LLL	O
)	O
Nédellec	O
et	O
al.77	O
sentences	O
from	O
research	O
articles	O
about	O
the	O
bacterium	O
Bacillus	O
subtilis	O
,	O
annotated	O
for	O
protein	O
–	O
gene	O
interactions.YesMedical	O
Subject	O
Headings	O
(	O
MeSH	O
)	O
National	O
Library	O
of	O
MedicineHierarchically-organized	O
terminology	O
for	O
indexing	O
and	O
cataloging	O
biomedical	O
documents.YesMetathesaurusNational	O
Library	O
of	O
Medicine	O
/	O
UMLS3.67	O
million	O
concepts	O
and	O
14	O
million	O
concept	O
names	O
,	O
mapped	O
between	O
more	O
than	O
200	O
sources	O
of	O
biomedical	O
vocabulary	O
and	O
identifiers.Yes	O
,	O
with	O
UMLS	O
License	O
AgreementMIMIC-IIIMIT	O
Lab	O
for	O
Computational	O
Physiologyde-identified	O
data	O
associated	O
with	O
53,423	O
distinct	O
hospital	O
admissions	O
for	O
adult	O
patients.Requires	O
training	O
and	O
formal	O
access	O
requestODIE	O
CorpusSavova	O
et	O
al.180	O
clinical	O
notes	O
annotated	O
with	O
5,992	O
coreference	O
pairs.NoOHSUMEDHersh	O
et	O
al.348	O
,	O
566	O
biomedical	O
research	O
abstracts	O
and	O
indexing	O
information	O
from	O
MEDLINE	O
,	O
including	O
MeSH	O
(	O
as	O
of	O
1991	O
)	O
.YesPMC	O
Open	O
Access	O
SubsetNational	O
Library	O
of	O
Medicine	O
/	O
PubMed	B-Library
CentralMore	O
than	O
2	O
million	O
research	O
articles	O
,	O
updated	O
weekly.YesRxNormNational	O
Library	O
of	O
Medicine	O
/	O
UMLSNormalized	O
names	O
for	O
clinical	O
drugs	O
and	O
drug	O
packs	O
,	O
with	O
combined	O
ingredients	O
,	O
strengths	O
,	O
and	O
form	O
,	O
and	O
assigned	O
types	O
from	O
the	O
Semantic	O
Network.Yes	O
,	O
with	O
UMLS	O
License	O
AgreementSemantic	O
NetworkNational	O
Library	O
of	O
Medicine	O
/	O
UMLSLists	O
of	O
133	O
semantic	O
types	O
and	O
54	O
semantic	O
relationships	O
covering	O
biomedical	O
concepts	O
and	O
vocabulary.Yes	O
,	O
with	O
UMLS	O
License	O
AgreementSPECIALIST	O
LexiconNational	O
Library	O
of	O
Medicine	O
/	O
UMLSA	O
syntactic	O
lexicon	B-Application
of	O
biomedical	O
and	O
general	O
English.YesWord	O
Sense	O
Disambiguation	O
(	O
WSD	O
)	O
National	O
Library	O
of	O
Medicine	O
/	O
UMLS203	O
ambiguous	O
words	O
and	O
37,888	O
automatically	O
extracted	O
instances	O
of	O
their	O
use	O
in	O
biomedical	O
research	O
publications	O
.	O
</s>
<s>
Several	O
groups	O
have	O
developed	O
sets	O
of	O
biomedical	O
vocabulary	O
mapped	O
to	O
vectors	O
of	O
real	O
numbers	O
,	O
known	O
as	O
word	B-General_Concept
vectors	I-General_Concept
or	I-General_Concept
word	I-General_Concept
embeddings	I-General_Concept
.	O
</s>
<s>
The	O
majority	O
are	O
results	O
of	O
the	O
word2vec	B-Algorithm
model	O
developed	O
by	O
Mikolov	O
et	O
al	O
or	O
variants	O
of	O
word2vec	B-Algorithm
.	O
</s>
<s>
+Biomedical	O
word	O
embeddingsSet	O
NameAuthors	O
or	O
GroupContents	O
and	O
SourceCitationBioASQword2vecBioASQVectors	O
produced	O
by	O
word2vec	B-Algorithm
from	O
10,876,004	O
English	O
PubMed	B-Library
abstracts.bio.nlplab.org	O
resourcesPyysalo	O
et	O
al.A	O
collection	O
of	O
word	B-General_Concept
vectors	I-General_Concept
produced	O
by	O
different	O
approaches	O
,	O
trained	O
on	O
text	O
from	O
PubMed	B-Library
and	O
PubMed	B-Library
Central.BioVecAsgari	O
and	O
MofradVectors	O
for	O
gene	O
and	O
protein	O
sequences	O
,	O
trained	O
using	O
Swiss-Prot.RadiologyReportEmbeddingBanerjee	O
et	O
al.Vectors	O
produced	O
by	O
word2vec	B-Algorithm
from	O
the	O
text	O
of	O
10,000	O
radiology	O
reports	O
.	O
</s>
<s>
Text	B-Algorithm
mining	I-Algorithm
applications	O
in	O
the	O
biomedical	O
field	O
include	O
computational	O
approaches	O
to	O
assist	O
with	O
studies	O
in	O
protein	O
docking	O
,	O
protein	O
interactions	O
,	O
and	O
protein-disease	O
associations	O
.	O
</s>
<s>
The	O
extraction	O
of	O
kinetic	O
parameters	O
from	O
text	O
or	O
the	O
subcellular	O
location	O
of	O
proteins	O
have	O
also	O
been	O
addressed	O
by	O
information	B-General_Concept
extraction	I-General_Concept
and	O
text	B-Algorithm
mining	I-Algorithm
technology	O
.	O
</s>
<s>
Text	B-Algorithm
mining	I-Algorithm
can	O
aid	O
in	O
gene	O
prioritization	O
,	O
or	O
identification	O
of	O
genes	O
most	O
likely	O
to	O
contribute	O
to	O
genetic	O
disease	O
.	O
</s>
<s>
One	O
group	O
compared	O
several	O
vocabularies	O
,	O
representations	O
and	O
ranking	B-Application
algorithms	I-Application
to	O
develop	O
gene	O
prioritization	O
benchmarks	O
.	O
</s>
<s>
An	O
agricultural	O
genomics	O
group	O
identified	O
genes	O
related	O
to	O
bovine	O
reproductive	O
traits	O
using	O
text	B-Algorithm
mining	I-Algorithm
,	O
among	O
other	O
approaches	O
.	O
</s>
<s>
Text	B-Algorithm
mining	I-Algorithm
enables	O
an	O
unbiased	O
evaluation	O
of	O
protein-disease	O
relationships	O
within	O
a	O
vast	O
quantity	O
of	O
unstructured	B-Application
textual	I-Application
data	I-Application
.	O
</s>
<s>
A	O
text	B-Algorithm
mining	I-Algorithm
study	O
assembled	O
a	O
collection	O
of	O
709	O
core	O
extracellular	O
matrix	O
proteins	O
and	O
associated	O
proteins	O
based	O
on	O
two	O
databases	O
:	O
MatrixDB	O
(	O
)	O
and	O
UniProt	O
.	O
</s>
<s>
This	O
set	O
of	O
proteins	O
had	O
a	O
manageable	O
size	O
and	O
a	O
rich	O
body	O
of	O
associated	O
information	O
,	O
making	O
it	O
a	O
suitable	O
for	O
the	O
application	O
of	O
text	B-Algorithm
mining	I-Algorithm
tools	O
.	O
</s>
<s>
They	O
used	O
a	O
phrase-mining	O
pipeline	O
,	O
Context-aware	O
Semantic	O
Online	B-Application
Analytical	I-Application
Processing	I-Application
(	O
CaseOLAP	O
)	O
,	O
then	O
semantically	O
scored	O
all	O
709	O
proteins	O
according	O
to	O
their	O
Integrity	O
,	O
Popularity	O
,	O
and	O
Distinctiveness	O
using	O
the	O
CaseOLAP	O
pipeline	O
.	O
</s>
<s>
The	O
text	B-Algorithm
mining	I-Algorithm
study	O
validated	O
existing	O
relationships	O
and	O
informed	O
previously	O
unrecognized	O
biological	O
processes	O
in	O
cardiovascular	O
pathophysiology	O
.	O
</s>
<s>
Search	O
engines	O
designed	O
to	O
retrieve	B-Application
biomedical	I-Application
literature	I-Application
relevant	O
to	O
a	O
user-provided	O
query	O
frequently	O
rely	O
upon	O
text	B-Algorithm
mining	I-Algorithm
approaches	O
.	O
</s>
<s>
Publicly	O
available	O
tools	O
specific	O
for	O
research	O
literature	O
include	O
PubMed	B-Library
search	O
,	O
Europe	B-Application
PubMed	I-Application
Central	I-Application
search	O
,	O
GeneView	O
,	O
and	O
APSE	O
Similarly	O
,	O
search	O
engines	O
and	O
indexing	O
systems	O
specific	O
for	O
biomedical	O
data	O
have	O
been	O
developed	O
,	O
including	O
DataMed	O
and	O
OmicsDI	O
.	O
</s>
<s>
Electronic	B-Application
medical	I-Application
records	I-Application
(	O
EMRs	O
)	O
and	O
electronic	B-Application
health	I-Application
records	I-Application
(	O
EHRs	B-Application
)	O
are	O
collected	O
by	O
clinical	O
staff	O
in	O
the	O
course	O
of	O
diagnosis	O
and	O
treatment	O
.	O
</s>
<s>
The	O
clinical	B-Application
Text	I-Application
Analysis	I-Application
and	I-Application
Knowledge	I-Application
Extraction	I-Application
System	I-Application
,	I-Application
or	I-Application
cTAKES	I-Application
,	O
annotates	O
clinical	O
text	O
using	O
a	O
dictionary	O
of	O
concepts	O
.	O
</s>
<s>
Computational	B-Architecture
frameworks	I-Architecture
have	O
been	O
developed	O
to	O
rapidly	O
build	O
tools	O
for	O
biomedical	B-General_Concept
text	I-General_Concept
mining	I-General_Concept
tasks	O
.	O
</s>
<s>
The	O
SparkText	O
framework	O
uses	O
Apache	B-Language
Spark	I-Language
data	O
streaming	O
,	O
a	O
NoSQL	B-General_Concept
database	O
,	O
and	O
basic	O
machine	O
learning	O
methods	O
to	O
build	O
predictive	B-General_Concept
models	I-General_Concept
from	O
scientific	O
articles	O
.	O
</s>
<s>
Some	O
biomedical	B-General_Concept
text	I-General_Concept
mining	I-General_Concept
and	O
natural	B-Language
language	I-Language
processing	I-Language
tools	O
are	O
available	O
through	O
application	B-Application
programming	I-Application
interfaces	I-Application
,	I-Application
or	I-Application
APIs	I-Application
.	O
</s>
<s>
NOBLE	O
Coder	O
performs	O
concept	O
recognition	O
through	O
an	O
API	B-Application
.	O
</s>
<s>
The	O
following	O
academic	O
conferences	O
and	O
workshops	O
host	O
discussions	O
and	O
presentations	O
in	O
biomedical	B-General_Concept
text	I-General_Concept
mining	I-General_Concept
advances	O
.	O
</s>
<s>
A	O
variety	O
of	O
academic	O
journals	O
publishing	O
manuscripts	O
on	O
biology	O
and	O
medicine	O
include	O
topics	O
in	O
text	B-Algorithm
mining	I-Algorithm
and	O
natural	B-Language
language	I-Language
processing	I-Language
software	O
.	O
</s>
