NER model

The NER model is a method for determining the accuracy of live subtitles on TV or at events that are created using speech recognition. The three letters stand for number, edition error recognition and error. It is an alternative to the traditional model WER (Word Error Rate, Word neatness ).

The NER model includes a formula for determining the quality of live subtitles: an NER value of 100 means that the content has been reproduced perfectly correct. To calculate the total number of words of live subtitle is taken and subtracted (caused by poor voice recognition), the editing and the recognition error. This number is divided by the total number of words of live subtitling and multiplied a hundred times.

It means

N ( number) = total number of words of live subtitling
E ( Edition error ) = -edit
R ( recognition error ) = recognition errors

In Switzerland, this measurement method is already used on public television. Other countries have also signaled interest.

The traditionally used WHO model, however, is static because it simply measures the literal deviation of what is said from what is written without taking into account that there may be edited live subtitles.

598066