Under Community Review

Hi Naoko,

As you may be aware, there's no provision in a translation memory to penalise a single segment so I don't quite see how we could implement such a feature as penalizing segments from a particular TM. The best solution I can imagine is that the segment is given an attribute to indicate which TM it came from and then, in SDL Trados Studio, you can apply a filter and set a penalty for those segments. Does this sound like a reasonable approach?



TM Sequence penalties should be preserved in project TM

In TMS, penalties can be applied to TMs in a TM Sequence to ensure reuse from the sequenced TMs is correctly prioritized. These penalties are not carried over into the SDL Trados project TM that is included in SDL Trados Synergy packages. Where multiple fuzzy matches for the same source segment from different TMs in the TM Sequence are included in the project TM this may lead to matches being presented to the translator in the wrong order, i.e. matches from penalized TMS TMs being shown with higher match values then matches from TMS TMs that are not penalized on the server. To ensure consistency between server-side processing and client-side presentation of TM matches, the penalties should be carried over into the project TM.