| Stron w wątku: < [1 2] | WF Pro: what are "fuzzy reps"? Autor wątku: pcs_MCIL
|
|---|
neilmac
Hiszpania
Local time: 11:32
hiszpański > angielski
+ ...
| Same difference | Mar 29, 2013 |
Stanislav Pokorny wrote:
Rolf Kern wrote:
fuzzy reps = more or less exact repetitions
fuzzy = similar repetitions
However, I do not use any CAT tools and am therefore not familiar with these funny distictions.
Rolf
I'm afraid this is not the case, Rolf.
- Fuzzy matches are less than 100% matches between a document and your current TM
- Fuzzy reps (or internal fuzzies, as Dominique writes) are fuzzy matches that only become fuzzies as you translate and build up your TM, i.e. fuzzy matches within a document/project.
This is more or less what I understand, but Rolf's idea is simpler. My point is that when an agency mentions them, it's usually to leverage your rates down...
| | | | Selcuk Akyuz 
Turcja
Local time: 13:32
angielski > turecki
+ ...
| short fuzzy matches without context | Mar 30, 2013 |
Short fuzzy matches - bad
(Fuzzy) matches without context - bad
Short fuzzy matches without context - unacceptable
Long exact matches without context - acceptable (but only as fuzzy matches for me)
Long exact matches with context - acceptable (better with 3-5 preceding/following segments)
What is a short fuzzy match? A segment shorter than 7-10 words. Even an exact (100%) match can be wrong (different subject, client specific terminology, gender, etc.). ... See more Short fuzzy matches - bad
(Fuzzy) matches without context - bad
Short fuzzy matches without context - unacceptable
Long exact matches without context - acceptable (but only as fuzzy matches for me)
Long exact matches with context - acceptable (better with 3-5 preceding/following segments)
What is a short fuzzy match? A segment shorter than 7-10 words. Even an exact (100%) match can be wrong (different subject, client specific terminology, gender, etc.).
http://www.proz.com/forum/cat_tools_technical_help/228971-a_cat_tool_for_translators_only.html
WFP does not support 101% (guaranteed/perfect/in-context/ice) matches but fuzzy repetitions are discussed here, a weird CAT tool!
Selcuk ▲ Collapse
| | | | | | Samuel Murray
Holandia
Local time: 11:32
Członek ProZ.com
od 2006
angielski > afrikaans
+ ...
Dominique Pivard wrote:
The truth is: most translators probably don't consider context matches a "must have" feature.
...
Even the ones about the most popular tools (memoQ and Studio) only got about 150 views.
I wonder how much of this is down to terminology. If I had heard that you had made videos about ICE matches or 101% matches, I would have visited the site, but if I had heard that the videos are about context matches, I would not have bothered. This is because I know of this feature by the first two names, and the third name sounds very similar to a completely different feature that one of my CAT tools has and that I'm not interested in (even though I know what "ICE" stands for).
I now watched the videos (except Heartsome and DVX) -- very nicely done.
One question (to Dominique): in the videos, only CafeTran proposed in-context matches during real-time translation, and in the other tools you used pre-translation for that. Does that mean the other tools don't offer in-context matching for real-time translation, or does it mean CafeTran doesn't offer pretranslation?
[Edited at 2013-03-30 12:12 GMT]
| | |
|
|
|
Samuel Murray
Holandia
Local time: 11:32
Członek ProZ.com
od 2006
angielski > afrikaans
+ ...
Selcuk Akyuz wrote:
Short fuzzy matches - bad
(Fuzzy) matches without context - bad
Short fuzzy matches without context - unacceptable
1. I think that most in-context matching systems are not sufficient for short fuzzy matches if all context segments are short. I have not seen an in-context matching system that takes into account more than one preceding and one following segment, and for multiple short segments that is not sufficient to ensure a good match.
2. I have managed to convince one of my WWC clients that short segments need to be charged at a higher rate (I explained it and the PM bought my explanation). So there is hope.
| | | | | Context matches in CafeTran | Mar 30, 2013 |
Samuel Murray wrote:
I now watched the videos (except Heartsome and DVX) -- very nicely done.
Thanks!
Samuel Murray wrote:
One question (to Dominique): in the videos, only CafeTran proposed in-context matches during real-time translation, and in the other tools you used pre-translation for that. Does that mean the other tools don't offer in-context matching for real-time translation, or does it mean CafeTran doesn't offer pretranslation?
Well, the thing is, context matches were only added in CafeTran after I had started that video series. In fact, the author (Igor Kmitowski) became aware of them after hearing about the series and added the feature because he thought it could be useful. If memory serves, context matches could only be obtained when translating "as you go", not via pretranslating. But that was the situation at the time, and things may be different now, because CafeTran is being developed at a very rapid pace. I'll leave it to people more familiar with CafeTran to give a more definite answer.
| | | | Selcuk Akyuz
Turcja
Local time: 13:32
angielski > turecki
+ ...
Samuel Murray wrote:
1. I think that most in-context matching systems are not sufficient for short fuzzy matches if all context segments are short. I have not seen an in-context matching system that takes into account more than one preceding and one following segment, and for multiple short segments that is not sufficient to ensure a good match.
I use 3 or 5 preceding/following exact matches in DVX. It can be configured by the user, what about 101 preceding/following exact matches? Yes, this is available in DVX. You can watch Dominique's video on DVX2.
2. I have managed to convince one of my WWC clients that short segments need to be charged at a higher rate (I explained it and the PM bought my explanation). So there is hope.
Hope most translators will do it, ask for higher rates when translating short segments (with or without a CAT tool).
| | | | | Context depth in Déjà Vu | Mar 31, 2013 |
Selcuk Akyuz wrote:
I use 3 or 5 preceding/following exact matches in DVX. It can be configured by the user, what about 101 preceding/following exact matches? Yes, this is available in DVX. You can watch Dominique's video on DVX2.
Context depth in indeed expressed in odd numbers in Déjà Vu: 3, 5, 7 etc. "3" means the actual segment, plus one segment before and one segment after. "5" means the actual segment, plus two segments before and two segments after. And so on.
| | | | | Stron w wątku: < [1 2] | To report site rules violations or get help, contact a site moderator: You can also contact site staff by submitting a support request » WF Pro: what are "fuzzy reps"? | Trados Business Manager Lite | Create customer quotes and invoices from within Trados Studio
Trados Business Manager Lite helps to simplify and speed up some of the daily tasks, such as invoicing and reporting, associated with running your freelance translation business.
More info » |
| | TM-Town | Manage your TMs and Terms ... and boost your translation business
Are you ready for something fresh in the industry? TM-Town is a unique new site for you -- the freelance translator -- to store, manage and share translation memories (TMs) and glossaries...and potentially meet new clients on the basis of your prior work.
More info » |
|
| | | | X Sign in to your ProZ.com account... | | | | | |
Login to reply/comment 
Translation news related to Wordfast
