The CAT Pane in Memsource displays the source text alongside matches from translation memories, term bases, and machine translations for each segment. It also displays any non-translatable matches.
- 101% - an in-context translation memory match. It is greater than a 100% match since the context also matches what is saved in the TM.
- 100% - an exact translation memory match.
- 78% is a fuzzy translation memory match (anything below 100%).
- MT stands for machine translation. (This will only be shown if it was previously enabled by a Project Manager.)
- TB is a suggested term from a term base.
- S is a subsegment match. These are somewhat similar to terms from a TB, except that they are automatically generated from the translation memory.
In addition to what's noted above, there are some additional features you might see:
- An asterisk on a term notifies users that a Note or Usage Information is available.
- The letters TB being greyed-out indicates a "New" term that has not yet been approved.
- A term in red term indicates a rejected term that should not be used during translation.
Non-translatable Matches
An underlined score of 100% or 99% can be a Non-Translatable match or an MTQE (Machine translation Quality Estimation) match. Hover your mouse over the underlined score to see the details. You can also see the details in the CAT pane when you click on the match there. Both 99% and 100% matches are likely to be segments that probably do not require translation but that should still be checked by a translator.
If the Display non-translatables in Memsource Editor option is enabled, segments identified as non-translatable are displayed in the Memsource Editor’s CAT pane. When an NT match is hovered over, a tooltip will indicate whether the match is AI or rule-based. This information is also visible at the bottom of the CAT pane when an NT match is selected.
MTQE Matches
The Memsource MTQE feature enables users to analyze a job and provides a percentage score for machine translation suggestions at the segment level before any post-editing is done. (In order to do this, MTQA must be enabled in the MT Engine settings.)
- 100%: Perfect MT output—probably no post-editing required
- 99%: Near-perfect MT output—possibly minor post-editing required, mostly for formatting or punctuation issues
- 75%: Good MT match, but likely to require some human post-editing
- No score: This means MTQE cannot confidently identify the quality (it may be high or low), so the output needs to be checked by a Linguist.
The Order of CAT Pane Results
The results given in the CAT pane will be displayed in the following order:
- Best translation memory or non-translateable match, based on the score.
- Term base items (If there is a preferred term, it will be shown on the top.)
- Translation memory, non-translateable, or MTQE. This is based on the scores.
- Machine translations (with no MTQE score)
- Sub-segment TM (marked by a pink “S”)
Displaying Differences for Fuzzy Matches
When a fuzzy match is selected in the CAT pane, the differences between the source segment and the fuzzy match get displayed.
Displaying Additional Information for Matches
When any match is selected in the CAT pane, the additional information (who translated the segment first, who edited it last, the TM, etc.) is displayed at the bottom window of the CAT pane. Further details are displayed if you hover your mouse over the window.
Displaying Additional Information for Terms
When a term is selected in the CAT pane, additional information is displayed if available. This includes any notes, usage examples and edit links (if the user is authorized to edit the term). Please be aware that only the Note and Usage Example of the target term are displayed.
