AI engineering glossary · Evaluation metrics
What is WER (word error rate)?
Word error rate (also called WER): The standard metric for speech recognition: the number of substituted, deleted and inserted words needed to turn the recognised text into the reference, divided by the number of reference words. Lower is better, and it can exceed 100% if there are many insertions.
Word error rate: an example
A 100-word reference with 6 substitutions, 3 deletions and 1 insertion gives WER = 10 / 100 = 10%.
Related terms
More Evaluation metrics terms
- Youden's J statistic
- Accuracy
- Accuracy paradox
- Adjusted Rand index
- AUC-ROC
- Average precision
- Balanced accuracy
- BERTScore
Where you learn it
Word error rate is part of the AI Engineering Program, which teaches it with notes, a hands-on lab and an exam. The first lesson is free.