Trados 2007 did come with MultiTerm. I can't remember if it was a separate installation disc or if it was on the same CD as Workbench.
Simone Duverger wrote:
But can anyone tell me how does it work for the less than 100% match segments? I thought if I have to translate a segment where for example one word is different, Trados will translate the other words directly when I open the tags to edit the segment?
In the "Options" menu in Workbench, choose "Translation Memory Options." The first item in the "General" tab is "Minimum match value." I'm guessing you have this set to a high number (maybe 95 or 100%). Drop it to 70% and you should start seeing fuzzy matches, assuming your TM contains segments that are 70%-99% matches for the segments you're trying to translate.
I tried to choose a terminological database in the menu "Options/Options for terminological recognition" but there is no mtw-file to add. I thought if I create some TM, I can automatically use it as terminology database for recognition? Or must I create a .mtw base before having the terminology recognition? How does it function?
TMs and termbases are two different things. TMs contain entire segments (usually sentences) and are useful for matched segments and concordance searches. But for translations of individual words (i.e., glossaries), you need to create a termbase in MultiTerm and then tell Workbench to use that particular termbase for Term Recognition.