Populating your TRADOS translation memory Thread poster: Todd Field
|
|---|
Todd Field 
United States
Local time: 00:41
Member
Portuguese to English
Dear fellow translators,
One more TRADOS question for a couple of rookies…
Aside from simply using the software repeatedly over time -- and excluding information you might obtain from an imported package -- are there other ways of populating your personal database with translation memories?
In this modern age, it seems that one might be able to import such information from another TRADOS user, or even purchase it based upon the an... See more Dear fellow translators,
One more TRADOS question for a couple of rookies…
Aside from simply using the software repeatedly over time -- and excluding information you might obtain from an imported package -- are there other ways of populating your personal database with translation memories?
In this modern age, it seems that one might be able to import such information from another TRADOS user, or even purchase it based upon the anticipated subject matter to be translated (i.e. engineering, finance, law, etc.).
Has anyone heard of creative ways in which this might be accomplished?
Thanks in advance for your responses,
Todd and Monica. ▲ Collapse
| | | |
Ralf Lemster
Germany
Local time: 08:41
English to German
+ ...
Hi Todd and Monica,
The most obvious source of material is to align projects done before you purchased Trados, using WinAlign. That process can be time-consuming: therefore, focus on documents that are representative of what you do, or have a certain probability of being repeated. (In the financial field, annual reports are the classic example for this.)
Buying translation memories is probably difficult, not because of technical problems, but because TM\'s represen... See more Hi Todd and Monica,
The most obvious source of material is to align projects done before you purchased Trados, using WinAlign. That process can be time-consuming: therefore, focus on documents that are representative of what you do, or have a certain probability of being repeated. (In the financial field, annual reports are the classic example for this.)
Buying translation memories is probably difficult, not because of technical problems, but because TM\'s represent a translator\'s (or agency\'s) asset, and determining the economic value of that asset is not exactly easy. There was a discussion on groups.yahoo.com/group/TW_users/ covering that topic recently. ▲ Collapse
| | | |
| Extract from online ressources.. | Mar 2, 2003 |
Extracting from online glossaries and dictionaries or collections is my way to create dictionaries for MultiTerm (Trados), TermWatch (Déjà Vu), or WordFast and MetaTexis.
I use Perl and/or Access to extract, clean and format such ressources. I have done it with the glossaries of Microsoft and SAP (about 10 MB each), but I may not sell or distribute them because of the copyright. Only a copy for personal use is permitted. 
[ This ... See more Extracting from online glossaries and dictionaries or collections is my way to create dictionaries for MultiTerm (Trados), TermWatch (Déjà Vu), or WordFast and MetaTexis.
I use Perl and/or Access to extract, clean and format such ressources. I have done it with the glossaries of Microsoft and SAP (about 10 MB each), but I may not sell or distribute them because of the copyright. Only a copy for personal use is permitted.
[ This Message was edited by:on2003-03-02 22:21] ▲ Collapse
| | |
|
|
|
| Extract from online ressources... (2) | Mar 4, 2003 |
Hi Harry,
Sorry to bother you personally, but this is a question I\'ve had since I bought Trados a month ago.
I\'ve tried to figure it out myself, but I\'m a bit of a techno-bimbo, I\'m afraid 
Could you spell out the steps you take to do what you have described? Let\'s say I have discovered a glossary on a website, of terms in the financial world, for example. I open the page, and then... how do I... See more Hi Harry,
Sorry to bother you personally, but this is a question I\'ve had since I bought Trados a month ago.
I\'ve tried to figure it out myself, but I\'m a bit of a techno-bimbo, I\'m afraid
Could you spell out the steps you take to do what you have described? Let\'s say I have discovered a glossary on a website, of terms in the financial world, for example. I open the page, and then... how do I proceed?
Thank you very much for your time and patience in this matter. You may even want to post the answer to the forum, as I am sure there are other individuals out there who need this vital information.
They really should add How to use Trados in the classic book line, XXX for Dummies!
Thanks again,
Nancy
Hi Nancy,
I already tried to explain it to non-programming colleagues,
but I did not succeed in explaining.
There are too many details which can - and use to - go wrong,
and even for experienced programmers it is always a long
procedure of trial and error.
For the extraction of glossaries you need to be skilled in:
A) a programming language like Visual Basic or Perl,
which can process text and open and read web sites,
B) creation and application of database systems,
like MS Access or any other SQL system (or at least
Excel, but it can handle only up to 65535 records)
Then you can extract and clean your term pairs
(because most databases are very dirty, i.e.
wrong content or wrong format in wrong field)
until you get a nice tab separated text file, which you
can directly import into Déjà Vu, WordFast or MetaTexis.
For Trados you need a further step of formatting
(embedding the text into Trados specific tags).
The challenge is to work faster than the results become old and useless.
For non-programmers I can only recommend to use the align facilities
of Trados or Déjà Vu, to create translation memories. But you need to
read the manual and exercise to get through with it.
To create terminology databases (as opposed to translation memories)
you would have to enter the single terms manually in Trados MultiTerm
or Déjà Vu TermWatch.
Sorry to have no better news,
Good luck, Harry
PS: in 2 weeks I will get an expertise from a lawyer,
about how I could distribute glossaries or help people to
create their own ones without violating any copyright
▲ Collapse
| | | |
| Microsoft Glossaries: Questions on legal implications | May 3, 2005 |
Dear Harry
Among the things, you wrote
PS: in 2 weeks I will get an expertise from a lawyer,
about how I could distribute glossaries or help people to
create their own ones without violating any copyright
I recently developed a procedure to generate X-glossaries, starting from both EN-L1 files and EN-L2 csv files. (L1,L2 = languages other tha English).
In case of French and Italian, this resulted in a huge L1-L2 csv file (containing approx. 39... See more Dear Harry
Among the things, you wrote
PS: in 2 weeks I will get an expertise from a lawyer,
about how I could distribute glossaries or help people to
create their own ones without violating any copyright
I recently developed a procedure to generate X-glossaries, starting from both EN-L1 files and EN-L2 csv files. (L1,L2 = languages other tha English).
In case of French and Italian, this resulted in a huge L1-L2 csv file (containing approx. 390K segments).
Now, my question. In your opinion, and based on your "expertise from a lawyer", publishing the procedure (in an article, for example) would infringe some copyright?
And, what about using Microsoft Glossaries (or a database derived from them) for translation purposes (when the Client is not MS Inc., of course)? Would this bother Bill?
Many thanks ▲ Collapse
| | | |
Maurizio Valente wrote:
Now, my question. In your opinion, and based on your "expertise from a lawyer", publishing the procedure (in an article, for example) would infringe some copyright?
And, what about using Microsoft Glossaries (or a database derived from them) for translation purposes (when the Client is not MS Inc., of course)? Would this bother Bill?
Dear Maurizio,
Publishing the procedure would not infringe any copyright, only publishing or providing the derived database would.
The advice from the lawyer was that the data should not be delivered via my web server. My server would only be allowed to act as a kind of "web dongle" for a program which users may use to download and process glossaries from third parties.
The point is, that I would not be obliged to ensure that those users would have a permission to do so.
Anyway, I did not develop programs of this kind any further, because I don't think there would be enough people really interested in buying them.
Concerning the use of MS Glossaries: According to MS terms and conditions, we are not even allowed to edit them. Nevertheless, I have not even heard of a single case where someone got into trouble by creating his own database derived from the MS Glossaries.
I think trouble could start when someone would advertise such a database, but I have never seen such an offer, neither in public nor in private.
In short, I don't see any problem with your intentions.
| | | |
| Many thanks ... | May 4, 2005 |
Many thanks, Harry, for your exhaustive and clear answer.
| | |
|
|
|
WordFast has recently released a beta version of a TM (translation memory) server it calls WfServer.
To use it, you download client software that connects to a WordSoft server through a TCP/IP socket. According to the manual, the server can access up to 2 million translation units per TM. Apparently it can access multiple TMs.
The only catch is that you need a license. Since I don't want to part with the 180 Euros, I haven't actually tried it out. If anyone
... See more WordFast has recently released a beta version of a TM (translation memory) server it calls WfServer.
To use it, you download client software that connects to a WordSoft server through a TCP/IP socket. According to the manual, the server can access up to 2 million translation units per TM. Apparently it can access multiple TMs.
The only catch is that you need a license. Since I don't want to part with the 180 Euros, I haven't actually tried it out. If anyone
has tried it, I'd be interested to read about the results. ▲ Collapse
| | | |
Login to reply/comment 
Translation news
