<s>
Text	B-General_Concept
simplification	I-General_Concept
is	O
an	O
operation	O
used	O
in	O
natural	B-Language
language	I-Language
processing	I-Language
to	O
change	O
,	O
enhance	O
,	O
classify	O
,	O
or	O
otherwise	O
process	O
an	O
existing	O
body	O
of	O
human-readable	O
text	O
so	O
its	O
grammar	O
and	O
structure	O
is	O
greatly	O
simplified	O
while	O
the	O
underlying	O
meaning	O
and	O
information	O
remain	O
the	O
same	O
.	O
</s>
<s>
Text	B-General_Concept
simplification	I-General_Concept
is	O
an	O
important	O
area	O
of	O
research	O
because	O
of	O
communication	O
needs	O
in	O
an	O
increasingly	O
complex	O
and	O
interconnected	O
world	O
more	O
dominated	O
by	O
science	O
,	O
technology	O
,	O
and	O
new	O
media	O
.	O
</s>
<s>
However	O
,	O
researchers	O
have	O
discovered	O
that	O
,	O
to	O
reduce	O
linguistic	O
diversity	O
,	O
they	O
can	O
use	O
methods	O
of	O
semantic	B-General_Concept
compression	I-General_Concept
to	O
limit	O
and	O
simplify	O
a	O
set	O
of	O
words	O
used	O
in	O
given	O
texts	O
.	O
</s>
<s>
Text	B-General_Concept
simplification	I-General_Concept
is	O
illustrated	O
with	O
an	O
example	O
used	O
by	O
Siddharthan	O
(	O
2006	O
)	O
.	O
</s>
<s>
A	O
text	B-General_Concept
simplification	I-General_Concept
system	O
aims	O
to	O
change	O
the	O
first	O
sentence	O
into	O
a	O
group	O
of	O
simpler	O
sentences	O
,	O
as	O
seen	O
just	O
below	O
the	O
first	O
sentence	O
.	O
</s>
<s>
One	O
approach	O
to	O
text	B-General_Concept
simplification	I-General_Concept
is	O
lexical	B-General_Concept
simplification	I-General_Concept
via	O
lexical	B-General_Concept
substitution	I-General_Concept
,	O
a	O
two-step	O
process	O
of	O
first	O
identifying	O
complex	O
words	O
and	O
then	O
replacing	O
them	O
with	O
simpler	O
synonyms	O
.	O
</s>
<s>
A	O
key	O
challenge	O
here	O
is	O
identifying	O
complex	O
words	O
,	O
which	O
is	O
performed	O
by	O
a	O
machine	O
learning	O
classifier	O
trained	O
on	O
labeled	B-General_Concept
data	I-General_Concept
.	O
</s>
