How many segments are usually included in a Translation Memory?
Thread poster: Masoud Kakoli

Masoud Kakoli
Iran
Local time: 07:51
English to Persian (Farsi)
+ ...
Apr 3, 2015

If you are familiar with CAT tools such as SDL, you probably know that this software turns every sentence into segment. If you have ever purchased a translation memory, how many segment were included in that TM? Or if you work with a TM of your own, how many segments are included in that TM?
If you have created your own TM from scratch, please let me know when you started to create it and how many segments are included. And if you purchased your TM, please let me know how many segments were included when you first purchased it and how much money you paid for it, if you are convenient with that.

[Edited at 2015-04-03 22:18 GMT]


Direct link Reply with quote
 

Natalie  Identity Verified
Poland
Local time: 06:21
Member (2002)
English to Russian
+ ...

Moderator of this forum
Misunderstanding Apr 3, 2015

Hi Masoud,

You have misunderstood the title of the courses: it is not about selling translation memories, it is "Selling Translations"TM, where TM stands for the "trademark" (i.e. it is Selling Translations®).

Purchasing somebody's TM hardly makes sense unless you work in some rather general field.

Natalia



[Edited at 2015-04-03 22:20 GMT]


Direct link Reply with quote
 

Masoud Kakoli
Iran
Local time: 07:51
English to Persian (Farsi)
+ ...
TOPIC STARTER
Have you ever purchased a TM? Apr 3, 2015

Natalie wrote:

Hi Masoud,

You have misunderstood the title of the courses: it is not about selling translation memories, it is "Selling Translations"TM, where TM stands for the "trademark" (i.e. it is Selling Translations®).

Natalia



Hi Natalie.
Thanks for your post. I edited the original question. Have you ever had such a experience? Do you have something to share with me?


Direct link Reply with quote
 

Natalie  Identity Verified
Poland
Local time: 06:21
Member (2002)
English to Russian
+ ...

Moderator of this forum
No, I use exclusively my own TMs Apr 3, 2015

And sometimes the TMs provided by my clients. The documents I work with are in 99% confidential, and most of my clients even prohibit using web-resources (all kinds of clouds etc.)

Moreover, when I use my own TMs I am sure in their quality.

I think there were discussions concerning TMs in the past - just make a search in the forums (CAT tools, SDL Trados, Translator resources etc.), you will certainly find something.


Direct link Reply with quote
 

Egidijus Slepetys  Identity Verified
Local time: 07:21
German to Lithuanian
New project - new TM Apr 4, 2015

After 10 years experience I can assure you - size doesn't matter. Even the largest TMs won't help. Big-mama-TMs are used only for looking-up for individual words or phrases, you can't remember how you translated them in the past.
New project - new TM. Only if a specific project is reoccurring, the correspondent TM comes into play.
Anything else doesn't work, because every project like every human is way too individual.
What works, is to have and with every project to fill you own termbase. But even terminology varies from project to project, and after years I'm changing some terms in it, because I don't like them anymore. Anyway, terminology is the best way to keep you confidently going, that's from my experience.
The termbase has to be filled by you - term for term with every project. If you import lists of terms, created by somebody else, that won't help much, because you won't feel confident.
The truth is - there is no easy way around.


Direct link Reply with quote
 

Masoud Kakoli
Iran
Local time: 07:51
English to Persian (Farsi)
+ ...
TOPIC STARTER
Concordance search Apr 4, 2015

Egidijus Slepetys wrote:

After 10 years experience I can assure you - size doesn't matter. Even the largest TMs won't help. Big-mama-TMs are used only for looking-up for individual words or phrases, you can't remember how you translated them in the past.
New project - new TM. Only if a specific project is reoccurring, the correspondent TM comes into play.
Anything else doesn't work, because every project like every human is way too individual.
What works, is to have and with every project to fill you own termbase. But even terminology varies from project to project, and after years I'm changing some terms in it, because I don't like them anymore. Anyway, terminology is the best way to keep you confidently going, that's from my experience.
The termbase has to be filled by you - term for term with every project. If you import lists of terms, created by somebody else, that won't help much, because you won't feel confident.
The truth is - there is no easy way around.


So you mean that the most important features of CAT tools are: the fact that these tools allow translators to use concordance search and the ability to use terminology management. Am I right?


Direct link Reply with quote
 

Erik Freitag  Identity Verified
Germany
Local time: 06:21
Member (2006)
Dutch to German
+ ...
Just the other way round Apr 4, 2015

Egidijus Slepetys wrote:

After 10 years experience I can assure you - size doesn't matter. Even the largest TMs won't help. Big-mama-TMs are used only for looking-up for individual words or phrases, you can't remember how you translated them in the past.
New project - new TM. Only if a specific project is reoccurring, the correspondent TM comes into play.
Anything else doesn't work, because every project like every human is way too individual.
What works, is to have and with every project to fill you own termbase. But even terminology varies from project to project, and after years I'm changing some terms in it, because I don't like them anymore. Anyway, terminology is the best way to keep you confidently going, that's from my experience.
The termbase has to be filled by you - term for term with every project. If you import lists of terms, created by somebody else, that won't help much, because you won't feel confident.
The truth is - there is no easy way around.


My experience is just the other way round: I don't find it very useful to separate TMs, so I generally follow a "Big Mama" approach (with a few exceptions). My termbases, however, are strictly separated for subject fields and/or clients.

If I have translated a sentence before, it will usually be useful for me to have a look at it when I get it a second time (even when in a different context). Of course, this is really a rare occurrence, but imho there really is no downside that would make the Big Mama approach unfavourable. Keeping separate TMs increases maintenance effort, though.

Terminology, however, is different: The chances that the same term appears in different contexts is quite high, and by keeping my termbases separate, I make sure that I only get relevant suggestions.

Your mileage may vary, of course.

[Bearbeitet am 2015-04-04 10:47 GMT]


Direct link Reply with quote
 
xxxbrg
Netherlands
Back to the original quesiton Apr 4, 2015

Masoud Kakoli wrote:

If you are familiar with CAT tools such as SDL, you probably know that this software turns every sentence into segment. If you have ever purchased a translation memory, how many segment were included in that TM? Or if you work with a TM of your own, how many segments are included in that TM?
If you have created your own TM from scratch, please let me know when you started to create it and how many segments are included. And if you purchased your TM, please let me know how many segments were included when you first purchased it and how much money you paid for it, if you are convenient with that.

[Edited at 2015-04-03 22:18 GMT]


TMs are not something you can buy or sell. They contain segments which are pieces of text owned by the client. Doing so is copyright infringement.
I have one TM per end client, created from scratch.
I started creating TMs as soon as I started using a CAT.
The biggest one contains 18,000 segments, for a client I work for since 10 years.
The other TMs contain between 100 and 6,000 segments.(stored in one folder and properly classified).
In Studio one can use several TMs at the same time, with their own read/write authorizations.
This is very useful to keep a special TM for a special client "clean" while having the possibility to have a look into the other TMs in the same field.
Because it is true that one of the best features of CATs is the concordance function.
Clearly speaking: one works on 1 TM at the same time, while the other TMs are there for memory purposes ('how did I translate XX a year ago for clients YY and ZZ?').
Personally I am not in favour of a big mamma approach, because clients generally have different styles and terminologies.
And creating a new TM for each project is totally inefficient.


Direct link Reply with quote
 

Heinrich Pesch  Identity Verified
Finland
Local time: 07:21
Member (2003)
Finnish to German
+ ...
Big TM good for concordance search Apr 5, 2015

It is true that you hardly get 100% matches from non-specific TMs, but for concordance search they work well. I do not use termbases, except when I have to.
For autosuggest TMs you need to have a minimum number of units.


Direct link Reply with quote
 

Meta Arkadia
Local time: 11:21
English to Indonesian
+ ...
It all depends... Apr 5, 2015

houtberg wrote:
And creating a new TM for each project is totally inefficient.

I create a new Project TM for every project. I save it in Dropbox, auto-save set to 1 segment, and saving tags. The very same segments of that project also go to my Big Mama, not saved in Dropbox, tags not saved.
Creating a TM is a matter of seconds, rather than minutes. Safety first!

Cheers,

Hans


Direct link Reply with quote
 

Meta Arkadia
Local time: 11:21
English to Indonesian
+ ...
But only... Apr 5, 2015

Heinrich Pesch wrote:
For autosuggest TMs you need to have a minimum number of units.

... if you still use Trados. I get autosuggest/autocomplete/autowhatever suggestions instantly from any attached TMs, and after the first segment also from the Project. The latter prevails, which makes sense.

Personally I am not in favour of a big mamma approach, because clients generally have different styles and terminologies.

They do, but you don't have to enable auto-assembly, just because you can. A Big Mama is terrific for concordance search - as you mentioned - and standard phrases. So why not use it?

I do not use termbases, except when I have to.

And to make things worse, I consider termbases (plural, for each project, with the possible exception of a completely new subject) and auto-assemble the core of my CAT tool use.

Well, each to his own.

Cheers,

Hans


Direct link Reply with quote
 

Meta Arkadia
Local time: 11:21
English to Indonesian
+ ...
By the way... Apr 5, 2015

and after the first segment also from the Project.

... this is also a reason (for me) to create a ProjectTM. It's obvious, that auto-complete suggestions from the Project are more likely to fit in the current translation than from other TMs. And yet another reason for creating a ProjectTM is, that I can set it - like any other TM - to "Terms consistency check", but with the highest priority, which also makes sense when you do your QA.

Cheers,

Hans


Direct link Reply with quote
 


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


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

How many segments are usually included in a Translation Memory?

Advanced search







WordFinder
The words you want Anywhere, Anytime

WordFinder is the market's fastest and easiest way of finding the right word, term, translation or synonym in one or more dictionaries. In our assortment you can choose among more than 120 dictionaries in 15 languages from leading publishers.

More info »
Déjà Vu X3
Try it, Love it

Find out why Déjà Vu is today the most flexible, customizable and user-friendly tool on the market. See the brand new features in action: *Completely redesigned user interface *Live Preview *Inline spell checking *Inline

More info »



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