<s>
The	O
NER	B-General_Concept
model	I-General_Concept
is	O
one	O
of	O
a	O
number	O
of	O
methods	O
for	O
determining	O
the	O
accuracy	O
of	O
live	O
subtitles	O
in	O
television	O
broadcasts	O
and	O
events	O
that	O
are	O
produced	O
using	O
speech	B-Application
recognition	I-Application
.	O
</s>
<s>
It	O
is	O
an	O
alternative	O
to	O
the	O
WER	B-General_Concept
model	I-General_Concept
(	O
Word	B-General_Concept
Error	I-General_Concept
Rate	I-General_Concept
)	O
used	O
in	O
several	O
countries	O
.	O
</s>
<s>
This	O
measurement	O
process	O
is	O
already	O
used	O
for	O
public	O
television	O
broadcasts	O
in	O
several	O
European	O
countries	O
like	O
Italy	O
and	O
Switzerland	B-Protocol
.	O
</s>
<s>
By	O
way	O
of	O
contrast	O
,	O
the	O
WER	B-General_Concept
model	I-General_Concept
is	O
static	O
,	O
since	O
it	O
simply	O
measures	O
the	O
textual	O
discrepancy	O
between	O
that	O
which	O
was	O
written	O
and	O
spoken	O
.	O
</s>
