The NER model is one of a number of methods for determining the
accuracy
Accuracy and precision are two measures of ''observational error''.
''Accuracy'' is how close a given set of measurements ( observations or readings) are to their ''true value'', while ''precision'' is how close the measurements are to each oth ...
of live
subtitles in
television
Television, sometimes shortened to TV, is a telecommunication medium for transmitting moving images and sound. The term can refer to a television set, or the medium of television transmission. Television is a mass medium for advertisin ...
broadcasts and events that are produced using
speech recognition
Speech recognition is an interdisciplinary subfield of computer science and computational linguistics that develops methodologies and technologies that enable the recognition and translation of spoken language into text by computers with the ...
. The three letters stand for number, edition error and recognition error. It is an alternative to the
WER model (Word Error Rate) used in several countries.
The model contains a
formula to determine the quality of live subtitles: a NER value of 100 indicates that the content was subtitled entirely correctly. This overall
score is calculated as follows: Firstly, the number of edit and recognition errors is deducted from the total number of words in the live subtitles. This number is then divided by the total number of words in the live subtitles and finally multiplied by one hundred.
:
.
The acronyms stand for the following:
* ''N'' (number) = total number of words in the live subtitles
* ''E'' (Edition error) = edition error
* ''R'' (Recognition error) = recognition error
This measurement process is already used for public television broadcasts in several European countries like
Italy
Italy ( it, Italia ), officially the Italian Republic, ) or the Republic of Italy, is a country in Southern Europe. It is located in the middle of the Mediterranean Sea, and its territory largely coincides with the homonymous geographical ...
and
Switzerland. Other countries and authorities like
British Ofcom have already expressed an interest.
By way of contrast, the WER model is static, since it simply measures the textual discrepancy between that which was written and spoken.
See also
*
Word error rate Word error rate (WER) is a common metric of the performance of a speech recognition or machine translation system.
The general difficulty of measuring performance lies in the fact that the recognized word sequence can have a different length from ...
References
* Pablo Romero-Fresco: Subtitling through Speech Recognition: Respeaking. Manchester: St. Jerome 2011, {{ISBN, 9781905763283
External links
presentation of the NER model at the International Telecommunication Union - Geneva, November 25, 2011
British Ofcom - The quality of live subtitling, May 2013
Speech recognition