<s>
Concept	B-General_Concept
mining	I-General_Concept
is	O
an	O
activity	O
that	O
results	O
in	O
the	O
extraction	O
of	O
concepts	O
from	O
artifacts	O
.	O
</s>
<s>
Solutions	O
to	O
the	O
task	O
typically	O
involve	O
aspects	O
of	O
artificial	B-Application
intelligence	I-Application
and	O
statistics	O
,	O
such	O
as	O
data	B-Application
mining	I-Application
and	O
text	B-Algorithm
mining	I-Algorithm
.	O
</s>
<s>
The	O
thesauri	O
used	O
are	O
either	O
specially	O
created	O
for	O
the	O
task	O
,	O
or	O
a	O
pre-existing	O
language	O
model	O
,	O
usually	O
related	O
to	O
Princeton	O
's	O
WordNet	B-General_Concept
.	O
</s>
<s>
Humans	O
use	O
context	O
to	O
disambiguate	B-General_Concept
the	O
various	O
meanings	O
of	O
a	O
given	O
piece	O
of	O
text	O
,	O
where	O
available	O
machine	B-Application
translation	I-Application
systems	I-Application
cannot	O
easily	O
infer	O
context	O
.	O
</s>
<s>
For	O
the	O
purposes	O
of	O
concept	B-General_Concept
mining	I-General_Concept
,	O
however	O
,	O
these	O
ambiguities	O
tend	O
to	O
be	O
less	O
important	O
than	O
they	O
are	O
with	O
machine	B-Application
translation	I-Application
,	O
for	O
in	O
large	O
documents	O
the	O
ambiguities	O
tend	O
to	O
even	O
out	O
,	O
much	O
as	O
is	O
the	O
case	O
with	O
text	B-Algorithm
mining	I-Algorithm
.	O
</s>
<s>
There	O
are	O
many	O
techniques	O
for	O
disambiguation	B-General_Concept
that	O
may	O
be	O
used	O
.	O
</s>
<s>
These	O
are	O
numerically	O
far	O
more	O
efficient	O
than	O
their	O
text	B-Algorithm
mining	I-Algorithm
cousins	O
,	O
and	O
tend	O
to	O
behave	O
more	O
intuitively	O
,	O
in	O
that	O
they	O
map	O
better	O
to	O
the	O
similarity	O
measures	O
a	O
human	O
would	O
generate	O
.	O
</s>
