Machine translation quality estimation
Automatically evaluate the quality of machine-translated text
Language Weaver's Machine translation Quality Estimation (MTQE) gives an overall prediction of the quality of machine translation output in real-time, helping to identify where translations may be inaccurate or unclear. This feedback can be used to identify high-quality translations that can be used as-is and low-quality translations that needs further review or editing by a professional translator.
Enable continuous improvement and refinement
Evaluating the quality of machine-translated text is crucial, as the output can vary based on factors such as the language pair, domain, and complexity of the input text.
Language Weaver’s built-in MT quality estimation is fast and scalable. It can evaluate text at both the sentence and document levels, depending on your needs, and it doesn't require human input during the evaluation process.
If you need automated decision-making and assistance in identifying MT output that may need human post-editing or revision, Language Weaver’s MT quality estimation tool is user-friendly and can be easily integrated into your workflows.
Gain efficiency in the Feedback Editor
Language Weaver's feedback editor option allows users to easily provide real-time, direct feedback to our MT output by suggesting better translations. Users can edit MT output for entire documents or individual segments at a time, directly and natively within our secure environment.
With feedback editor, users can see Language Weaver’s quality estimation at the segment level, thereby knowing where to focus their post-editing efforts.
Full integration with workflows
Plug and play with Trados enables businesses to take advantage of quality estimation in their localization workflows.
