What is your experience with keeping multiple translation memories in Trados?
Thread poster: motuzik

motuzik
Slovakia
Local time: 13:03
May 13, 2009

Hi all,

What is your experience with keeping multiple translation memories in Trados? I am wondering whether the system is more convenient to be used with a single bulk memory where you store all translated units, or should I create a separate memory for each project?? And what would be the advantage of using separate memories?

Thanks

Matus


[Subject edited by staff or moderator 2009-05-13 16:31 GMT]


 

S_Angel  Identity Verified
Italy
Local time: 13:03
German to Italian
+ ...
Different TMs May 13, 2009

Personally, I prefer to use a different TM for each client, since each one may have different terminology preferences. I only use the same TM when the subject is the same, or very similar.
But I guess it is just a matter of personal preference.icon_smile.gif


 

Stanislav Pokorny  Identity Verified
Czech Republic
Local time: 13:03
English to Czech
+ ...
Organizing TMs May 13, 2009

Hi Matúš,
I usually use different TMs for different clients and sometimes even for different client job. I use the following directory structure: Documents and Settings/My account name/Documents/Translations/TMs/Client name/tm.tmw. The TM name reflects the client name, topic, TM version and language pair (e.g. client_technical_21_engb_cscz.tmw).


 

Adam Łobatiuk  Identity Verified
Poland
Local time: 13:03
Member (2009)
English to Polish
+ ...
Keep them separate May 13, 2009

Hi Matus,

Imagine working for 2 end-clients from the same sector who require different terminology. You'll end up with a terminological mess if you use a single large TM. Apart from that, your clients might expect different style, register (e.g. how you address the reader) etc. In fact, they might expect different style in different types of content, like web pages and user guides.

However, you might try merging exported TM's and using that to translate exported "unknown segments". You might gain some translations that way, but of course they will require editing.


 

Clarisa Moraña  Identity Verified
Argentina
Local time: 08:03
Member (2002)
English to Spanish
+ ...
one big memory May 13, 2009

And each client is differenciated with filters; filters are an excellente tool!
Sure, I do have several translation memories (localization, oil & gas, general, timber and medicine), but each one of those contains filters for each client (and their preferred terminology). I consider that one big memory is the best way to leverage all my previous work.

Kind regards

Clarisa Moraña


 

Giles Watson  Identity Verified
Italy
Local time: 13:03
Italian to English
YMMV May 13, 2009

Hi Matus,

Your mileage may vary, as they say.

The crucial point here is to analyse your client portfolio and then decide what approach to adopt.

The options range from Clarisa's "Big Momma" approach (as DV users used to, and probably still do, call the single-TM strategy) to individual TMs for every clients, or even project.

I find it more convenient to use client-specific TMs combined into larger sector-wide TMs that I can then use as reference memories. Sadly, Trados has a rather weak concordance feature, which limits the number of instances you can call up in a single search to 99 and, depending on what you are looking for, may exclude the most relevant hits if you are using a large TM. To circumvent this limitation, I often use the freeware ApSIC concordancer in tandem with Workbench.

But whatever your final strategy, you will have to find your own trade-off between the completeness of your search and the "noise" built in by partially relevant TMs.

Best,

Giles


 

Tomás Cano Binder, BA, CT  Identity Verified
Spain
Local time: 13:03
Member (2005)
English to Spanish
+ ...
By end customer May 13, 2009

Although I agree that the ideal approach would be a big memory and use of filters, in the rush of things the risk of forgetting to configure the filters every time you switch jobs is rather high.

The consequences of reusing a segment from a competitor (and the risk of not noticing that a product, person, company or trademark name from a competitor could end up in your job) can be catastrophic and could go far beyond a reprimand from the customer, so my recommendation is to keep one TM per client or, if you work for agencies who send you jobs from half a dozen different end customers, one per end end customer.

If the potential advantage of having one big memory is to have a record of terminology you can reuse in the long run (I don't see big chances of reuse from one customer to the next unless many of your customers work in the same regulated industry), but for terminology-tracking purposes you can use Multiterm and create a nice termbase/glossary.


 

Andrea Riffo  Identity Verified
Chile
Local time: 07:03
English to Spanish
Agree May 13, 2009

Tomás Cano Binder, CT wrote:

The consequences of reusing a segment from a competitor (and the risk of not noticing that a product, person, company or trademark name from a competitor could end up in your job) can be catastrophic and could go far beyond a reprimand from the customer, ...



This is exactly why I keep different TMs for different end clients.

Greetings
Andrea


 

Astrid Elke Witte  Identity Verified
Germany
Local time: 13:03
Member (2002)
German to English
+ ...
There is the invoicing aspect May 14, 2009

I use one big TM, as I find that the most useful in leveraging past translation units and saving time.

Each time I start a new project I put in the project number, so that I can at any time export all TUs containing that project number.

I keep a separate TM for each company as well - this may either mean for each end client of a firm of lawyers, for one of my own direct clients, or for everything from one agency if that agency (like most agencies with whom I have any dealings) only sends me jobs infrequently.

The big TM is - out of necessity in my language pair - set to accept multiple translations. This, of course, makes quality control of the TUs considerably more difficult - or rather impossible (the TU containing a mistake is saved in the TM, as well as the corrected version).

Once I have my final version of a translation, have proofread it on paper as well, and am convinced that the TUs for that project are exactly right, I export the TUs for that project into the individual company TM.

The purpose of this exercise is not to give a company a discount for too many matches.

I give full discounts of my own accord for exact matches, and partial discounts for 95-99% matches and 85-94% matches. In particular in regard to the exact matches, this is because lawyers often copy and paste huge chunks - which they are quoting - from previous letters, and it is unethical to charge for the same large chunk of text twice. However, they sometimes also send me several drafts in a series, and this can create a lot of high similarity matches.

Since, however, there is also the aspect - in another situation, e.g. when translating contracts - that identical or nearly identical chunks of text will occur when the documents come from totally different companies, if I run an analysis based on the TM for that company I will be excluding any exact matches created by translating that chunk of text for a different company. This is to make sure that I benefit - rather than a new customer, unjustifiably - when the same clause of a contract turns up.

So, in other words, I always run the analysis against the specific company TM (in order not to inadvertently charge too little), but translate in the very large TM in order to use past translations with maximum effect.

Astrid


 

motuzik
Slovakia
Local time: 13:03
TOPIC STARTER
Thank you May 14, 2009

Dear all,

thanks for this great variety of points of view, I think Astrid has put it most comprehensively.

I can already see the light at the end of the tunnel.icon_wink.gif

thanks all

Matus


 


To report site rules violations or get help, contact a site moderator:


You can also contact site staff by submitting a support request »

What is your experience with keeping multiple translation memories in Trados?

Advanced search







SDL MultiTerm 2017
Guarantee a unified, consistent and high-quality translation with terminology software by the industry leaders.

SDL MultiTerm 2017 allows translators to create one central location to store and manage multilingual terminology, and with SDL MultiTerm Extract 2017 you can automatically create term lists from your existing documentation to save time.

More info »
Protemos translation business management system
Create your account in minutes, and start working! 3-month trial for agencies, and free for freelancers!

The system lets you keep client/vendor database, with contacts and rates, manage projects and assign jobs to vendors, issue invoices, track payments, store and manage project files, generate business reports on turnover profit per client/manager etc.

More info »



Forums
  • All of ProZ.com
  • Term search
  • Jobs
  • Forums
  • Multiple search