Machine Translation Detection
Proprietary detection.
Our proprietary MT Detection technology uniquely validates translation naturality by detecting possible use of machine translation (MT) and/or post-editing (MTPE) in translated strings.
Built in
e2f’s MT Detection technology is a proprietary tool built into its translation workflow platform.
Live comparison
As translators post their translations to the platform, e2f’s MT Detection tool automatically compares the human-supplied translations with translations generated by nearly all MT engines available to translators.
String matching
Using a proprietary algorithm, e2f’s tool can determine whether the human’s string is an exact match of one of the machine-translated strings, or if it closely resembles one of the MT strings.
Human validation
Suspicious strings are flagged for human review. A human reviewer will then decide whether the human’s work is really the way a human would say the thing, or if it looks a bit unnatural and should be retranslated.
Clean strings
e2f validates that every translated string is free of any trace of MT and MTPE.