Importing MS glossaries into Trados Thread poster: robroy
|
|---|
robroy
Local time: 19:02
German to English
I have just downloaded the German MS glossary, and it is a mind-boggling 850 MB with 234 csv files.
My question - is it better to import them into the Workbench or into MultiTerm? In both cases the databases will be absolutely enormous, presumably slowing down my system drastically.
The advantage in MultiTerm would be that in the latest version (MT 7) it is possible to open multiple termbases. In Workbench 7 it is still only possible to open one memory plus a second as concordance r... See more I have just downloaded the German MS glossary, and it is a mind-boggling 850 MB with 234 csv files.
My question - is it better to import them into the Workbench or into MultiTerm? In both cases the databases will be absolutely enormous, presumably slowing down my system drastically.
The advantage in MultiTerm would be that in the latest version (MT 7) it is possible to open multiple termbases. In Workbench 7 it is still only possible to open one memory plus a second as concordance reference.
Does anyone have any experience in working with the MS glossaries in Trados?
Windows XP Home
MultiTerm 7
Workbench 7 ▲ Collapse
| | | |
JennyC08 (X)
Local time: 13:02
German to French
+ ...
Hi,
Personnaly, I would prefer to import the glossary in the Workbench. It allows the Concordance search that you cannot use in the Multiterm (the phrases are classified by alphabetical order).
If you carry out a search on the forum, you will find some threads related to it.
HTH
Caroline
| | | |
| |
| Winlexic 2005 | Nov 1, 2005 |
Hi,
I've recently downloaded the trial version of Winlexic 2005 and I like what I've seen so far. Check out: http://www.winlexic.com
Cheers,
Evelyn
| | |
|
|
|
Zhoudan 
Local time: 01:02
English to Chinese
+ ...
| The latest MS glossaries are not supported | Nov 1, 2005 |
I use this freeware very often, but I'm unable to import the latest MS glossaries into this Xbench. Did I miss anything? Or is there an updated Xbench version?
| | | |
Zhoudan
Local time: 01:02
English to Chinese
+ ...
I tried the latest Xbench and it worked!
Zhoudan wrote: I use this freeware very often, but I'm unable to import the latest MS glossaries into this Xbench. Did I miss anything? Or is there an updated Xbench version?
| | | |
robroy
Local time: 19:02
German to English
TOPIC STARTER | Importing MS glossaries into Trados | Nov 3, 2005 |
Thank you Caroline, Laurent, Evelyn and Zhoudan.
Caroline - my idea was to use my standard memory in the Workbench, possibly with the Microsoft memory in the background, but then I can only use the Microsoft memory for concordance and not for finding actual matches. My idea was perhaps to import into MultiTerm, and then any terminological (or also phrase) matches are shown immediately, I don't have to start looking for them.
Laurent - I have tried XBench and some simila... See more Thank you Caroline, Laurent, Evelyn and Zhoudan.
Caroline - my idea was to use my standard memory in the Workbench, possibly with the Microsoft memory in the background, but then I can only use the Microsoft memory for concordance and not for finding actual matches. My idea was perhaps to import into MultiTerm, and then any terminological (or also phrase) matches are shown immediately, I don't have to start looking for them.
Laurent - I have tried XBench and some similar programs. But - see above - I still have to go out looking for the termini, whereas in MT they would appear automatically. Another point I already mentioned too - I tried to import all 234 files of the German MS glossary into Xbench, and the program basically died on me. As I say, the sheer size of the glossaries is a problem. Or does anyone know how much redundancy there is in these files, and if it is enough to use, say, the file for OfficeXP?
Evelyn - thank you, I have also heard of WinLexic, it appears to do the same as XBench, I don't know if it is better able to cope with the huge MS files. Any experience so far?
Zhoudan - glad that XBench is now working for you. What size of file did you import?
Summing up - it looks as if no one has imported the enormous MS files into MultiTerm, but instead the alternative search-and-find programs seem to be the best way.
If anyone has imported these files into Trados, was there much performance loss? What hardware do you use? (CPU, RAM ...)
Thanks again for your help! ▲ Collapse
| | | |
Vito Smolej
Germany
Local time: 19:02
Member (2004)
English to Slovenian
+ ...
SITE LOCALIZER | I made a translation memory out of it... | Nov 4, 2005 |
My question - is it better to import them into the Workbench or into MultiTerm?
As the glossary contains sentences and not words or idioms, the natural place to have is a TM. If you just stick to [EN - your target] pair, it's still some hard work, before you have it pat. For instance, you can't use excel, even if it is tempting: the bigger and more important CSVs are just too big to fit into 65000 lines limit.You can use word to process it, but it is still gargantuan.
Regards
smo
PS: It would be nice if TWB would allow (like Wordfast) several TMs to coexist - and then one could do concordance search in Gloassry TM etc -.
PPS: I may try - for the h*l of it - to make MultiTerm MDB out of it. Stand by for further news;)
[Edited at 2005-11-04 07:06]
| | |
|
|
|
pep
Local time: 19:02
English to Spanish
| Xbench clarifications | Nov 4, 2005 |
robroy wrote:
Laurent - I have tried XBench and some similar programs. But - see above - I still have to go out looking for the termini, whereas in MT they would appear automatically. Another point I already mentioned too - I tried to import all 234 files of the German MS glossary into Xbench, and the program basically died on me. As I say, the sheer size of the glossaries is a problem. Or does anyone know how much redundancy there is in these files, and if it is enough to use, say, the file for OfficeXP?
We use Xbench a lot (because we wrote it) and since it is not an indexer, you must be a little bit selective on what you load in the current version published. What we normally do in real projects is to load one of the Windows glossaries (for example, Windows XP if we are translating for end users products, Windows 2003 if we are translating for server software and when it makes sense, possibly a few of other more specialized glossaries (for example, SQL Server if the content we have to translate deals with databases). We are little bit picky on the terminology we add to Xbench for each project to avoid the problems derived from an excess of reference when you work with several translators in the same project.
On the other hand, regarding the redundancy, I know for a fact that there is a lot of redundancy between Windows 2003 and Windows XP as both share a lot of strings, so IMHO it does not make much sense to have them loaded at the same time.
If you have an average workstation machine, with the current version of Xbench, I don't advise to go beyond references with more than 1-2 million segments (that should be some 10-20 million words).
If you prefer to have all the terminology available regardless of the field, one alternative with Xbench is to load them in batches. Just sort the files per size and try to load some 100-200 MB in each batch and export them as TMX. In a few batches (I'm say something between 5 or 10) you should have a few TMX files that you could load into a Trados memory or some other indexed source.
Regarding using MultiTerm iX as a repository for all Microsoft glossaries, I'm not 100% sure because we use Xbench instead, but I suspect there will be performance issues given the underlying database technology. Probably your best bet for such a large corpus is Trados concordance, which means that what you should do is load the TMX files from Xbench (or from other tools that similar conversions) into a Trados memory and use it in the background.
Hope this clarifies a little bit how Xbench is currently implemented,
Pep.
| | | |
Login to reply/comment 
Translation news related to SDL Trados
