When you open a segment for translation, Matecat will automatically pre-populate it with the best TM or MT match available.
By design, if the machine translation is not enabled, Matecat populates the segments only if it detects a match from the TM, which is higher than 70%, otherwise, it will leave the target segment blank.
When the machine translation is enabled, Matecat populates the segments with matches from the TM, only if they are higher than 85% otherwise, it will use MT suggestions.
Translation Matches tab - This tab shows you the three highest matches available for the current segment, including the match percentage, its source and its creation date. Use the shortcut CTRL+1/2/3 to choose and edit one of the options.
You will receive TM suggestions if you add a private TM key to the project or fuzzy matches found in the public translation memory, as well as MT suggestions if activated.
The suggestions are ranked according to their match percentage, from the highest to the lowest. The MT suggestions are allocated an 85% match by default. As previously mentioned, in the absence of higher matching percentages, the MT will be displayed in the first place.
For every TM suggestion, its source, match percentage and creation date are indicated in the corner of every entry, as in the screenshot below:
Example includes:a 98% match from the private TM key added to the project (the private TM key is shown as the source of the suggestion) as the first suggestion, and therefore used to automatically populate the segment;
the Machine Translation suggestion (with the corresponding source label MT)
a 67% match from the public TM (source is “anonymous” here) as the third suggestion;
Depending on the source of the match and the match percentage, the following labels can be displayed:
101% (Ice) matches from the Translation Memory
100% matches from the Translation Memory
Fuzzy matches from the Translation Memory.
Machine Translation suggestions
There are two ways of selecting and inserting the preferred suggestion into the target segment:
- use the shortcut CTRL+[1/2/3] depending on the corresponding suggestion you have chosen, or
- double-click on the match of your choice.
To remove a TM match from the translation memory, hover over it and click on the dustbin icon
on the right.
NOTE: Use the removal tool only if you want to delete the match from the translation memory because the target text is not the correct translation of the source text. Do not delete it if you think it is not the correct suggestion for your current segment.