Kantan TotalRecall is automatically built into all translation operations on KantanMT. If a segment you are trying to translate is found in your data at a match of 85% or above, KantanMT will automatically leverage this segment from your training data.
TotalRecall is always working away in the background of your translations, but if you want to check how many words are being leveraged from your TM and how many are machine translated you can check by running an Analyse job.
- Upload your files for translation to the Translation tab.
- Click the Analyse button.
- When the job is complete, you can download the files and analysis from Jobs.
- To view Total Recall, click on the graph icon beside the download icon in Jobs.
- This takes you to the Analysis screen for your job. Here you can see the percentage Total Recall as well as the quality estimation score for all machine translated segments.
For more details on KantanTotalRecall please click on the link below: