Wouldn't it be nice for MemoQ to display fuzzy matches for termbase terms?
Chun-yi Chen United States Local time: 12:00 Member (2006) English to Chinese
TOPIC STARTER
custom settings
Dec 23, 2011
Hi Dominique,
I know a little bit of pipes and *'s through one of the MemoQ's recorded webinars, but I feel this is not very efficient, particularly if you are working with a termbase client sent you. Do you actually go in and change the inflected words (I didn't look up the word inflect, so forgive me if I use it incorrectly) one by one so that you get better match results?
Personally I think |'s and *'s are a workaround, not a solution. On the other hand, I have to say that since MemoQ is so smart in finding matches, I am no longer so sure if I do want to get "fuzzy" matches from the MemoQ termbases. It would certainly overcrowd the translation results panel and I don't want that, either;)
Chun-yi
a happy MemoQ user
Subject:
Comment:
The contents of this post will automatically be included in the ticket generated. Please add any additional comments or explanation (optional)
Marinus Vesseur Canada Local time: 12:00 Partial member English to Dutch + ...
Termbase handling "on the fly" better in SDL Studio
Dec 23, 2011
Hi Grigori
Grigori Gazarian wrote:
Marinus Vesseur wrote:
Yes, of course Studio's terminology is more comprehensive with Multiterm, but try to add ...
..
Sorry, but you misquoted me. That was Tomás' opinion.
I agree with you that Studio does a better job integrating the adding and editing of terms into termbases on the fly (with the usual fragile GUI, of course - nobody's perfect, especially not the nerdy makers of CAT tools), but, more importantly, it's easier to add them when you need to.
Mind you, I only recently found out that you can edit words in a MemoQ termbase, so I might be biased simply because I had Studio first.
What really bothers me with MemoQ is the long list of hits (and it keeps getting longer for several seconds after I jump into a segment) in all the colours of the rainbow, without a readily accessible explanation as to what these colours MEAN. Even if you find the explanation it doesn't help you much, because: what reliable value do those terms have in relation to other terms offered in the current context? I guess the sequence top-to-down has some meaning, but there is no context or pop-up menu that sheds any light on it.
Some terms stem "from fragments", some from a TD, some from a TM, some from "LSC hits", etc. Firstly, what do these abbreviations mean and secondly, why would I be helped by a list with 20 terms that look very similar or are the same? It all gives the impression of a collection of good ideas put together but not brought into harmony with one another.
The more I work with MemoQ the more I get the impression that it is only getting bigger, not better.
So much to that.
[Edited at 2011-12-24 04:25 GMT]
Subject:
Comment:
The contents of this post will automatically be included in the ticket generated. Please add any additional comments or explanation (optional)
Pipes and asterisks are optimal with terminology created by you
Dec 23, 2011
Chun-yi Chen wrote:
I know a little bit of pipes and *'s through one of the MemoQ's recorded webinars, but I feel this is not very efficient, particularly if you are working with a termbase client sent you. Do you actually go in and change the inflected words (I didn't look up the word inflect, so forgive me if I use it incorrectly) one by one so that you get better match results?
Personally I think |'s and *'s are a workaround, not a solution. On the other hand, I have to say that since MemoQ is so smart in finding matches, I am no longer so sure if I do want to get "fuzzy" matches from the MemoQ termbases. It would certainly overcrowd the translation results panel and I don't want that, either;)
You're right: it's time-consuming to add pipes and asterisks in existing terminology supplied by a client. I'm fortunate enough not to have this problem: all the terminology I use for my clients has been created from scratch by myself. When adding terminology as you translate, including pipes & asterisks (if needed) hardly requires extra effort.
Subject:
Comment:
The contents of this post will automatically be included in the ticket generated. Please add any additional comments or explanation (optional)
Chun-yi Chen United States Local time: 12:00 Member (2006) English to Chinese
TOPIC STARTER
I agree, the more are not necessarily the merrier
Dec 24, 2011
I have the same trouble as you described, Marinus.
It's sometimes counter-productive to scan through a long list of matches in the MemoQ translation results window. There are cases when I get four identical translation from the same text that has different formatting (some in bold, some in italics, that kind of stuff). It could push the ones I really wanted to use further down on the list and make them not readily be seen.
I also found that I receive the yellow subsegment hits (matches) which are actually empty, useless but take up those precious spaces in the results window.
There may be settings I could change to optimize the search results in MemoQ, but I haven't figured out where and how.
Marinus, have you tried to double click on the eye icon next to where the text Translation results is and change the Hit list filter settings? For example, change Filter and limit translation memory and corpus hit shown from the default 5 to 3. That may help decrease the number of hits listed on the results window.
Despite this inconvenience, I like MemoQ better than Studio. At least MemoQ gives me peace of mind, opens a project quickly and has hardly had any problems so far.
Chun-yi
Marinus Vesseur wrote:
What really bothers me with MemoQ is the long list of hits (and it keeps getting longer for several seconds after I jump into a segment) with all the colours of the rainbow, without a readily accessible explanation as to what these colours MEAN. Even if you find the explanation, it remains unclear what that means and what reliable value those terms have in relation to other terms offered. I guess the sequence top-to-down has some meaning, but there is no context or pop-up menu that helps you much.
Some terms stem "from fragments", some from a TD, some from a TM, some from "LSC hits", etc. Firstly, what does that mean and secondly, why would I be helped by a list with 20 terms that look very similar or are the same? It all gives the impression of a collection of good ideas put together but not brought into harmony with one another.
The more I work with MemoQ the more I get the impression that it is only get bigger, not better.
So much to that.
Subject:
Comment:
The contents of this post will automatically be included in the ticket generated. Please add any additional comments or explanation (optional)
Marinus Vesseur wrote:
Some terms stem "from fragments", some from a TD, some from a TM, some from "LSC hits", etc. Firstly, what do these abbreviations mean and secondly, why would I be helped by a list with 20 terms that look very similar or are the same? It all gives the impression of a collection of good ideas put together but not brought into harmony with one another.
LSC means "longest substring concordance". Based on your screenshot, it looks like you have misconfigured it (in Options > Subsegment leverage), aiming for quantity instead of quality. What do you have for minimum length in words/characters?
Subject:
Comment:
The contents of this post will automatically be included in the ticket generated. Please add any additional comments or explanation (optional)
Marinus Vesseur Canada Local time: 12:00 Partial member English to Dutch + ...
Don't remember configuring anything..
Dec 24, 2011
.. but at least I now know how to switch it off. This might actually be a useful feature under certain circumstances, but here again, even the "explanation" in MemoQ isn't helpful. I have a basic resentment against programmers who have forgotten what it is like not to be a programmer. What's the use of the best feature when you aren't making very clear to the majority of users when and how to use it? Keep putting yourselves in the shoes of your client base, MemoQ, or you might lose them.
Thank you for your help. Happy holidays! Hopefully you can afford not to have to work in them
Dominique Pivard wrote:
Marinus Vesseur wrote:
Some terms stem "from fragments", some from a TD, some from a TM, some from "LSC hits", etc...
LSC means "longest substring concordance". Based on your screenshot, it looks like you have misconfigured it (in Options > Subsegment leverage), aiming for quantity instead of quality. What do you have for minimum length in words/characters?
Subject:
Comment:
The contents of this post will automatically be included in the ticket generated. Please add any additional comments or explanation (optional)
Marinus Vesseur wrote:
Don't remember configuring anything .... but at least I now know how to switch it off.
The shortest orange hit in your screenshot was a single word of four characters (nach). This means you (or someone else) must have changed the default, which is 3 words and 16 characters. These parameters may need to be adjusted depending on your source language and on the size of your TM, but you definitely want to have more than one word and ten characters.
Marinus Vesseur wrote:
This might actually be a useful feature under certain circumstances, but here again, even the "explanation" in MemoQ isn't helpful. I have a basic resentment against programmers who have forgotten what it is like not to be a programmer.
I don't think you need to be a programmer in order to understand how LSC works, especially if you're familiar with other CAT tools: it's just a concordance search performed automatically for you, showing everything longer (in number of words and characters) than what you have specified. The default thresholds (3 words, 16 characters) probably work well with most languages and TM's of normal size.
Subject:
Comment:
The contents of this post will automatically be included in the ticket generated. Please add any additional comments or explanation (optional)
Marinus Vesseur Canada Local time: 12:00 Partial member English to Dutch + ...
Default MemoQ LSC settings do not seem to work
Dec 25, 2011
Dominique Pivard wrote:
..This means you (or someone else) must have changed the default.
No, it does not mean that at all. This is what I find when I look up those settings:
According to your description these are the default settings. Of course, it's impossible that a program like MemoQ should ever do anything wrong. It MUST be the stupid user. So I must have "done something wrong" elsewhere to bring this about that. Me, my ignorant self and I will keep looking for it, but in the meantime I'll switch that function off.
I don't think you need to be a programmer in order to understand how LSC works,..
Sorry, but that, too, is somewhat of an arrogant assumption. I know that you didn't consciously mean to imply it, but I never said it would be impossible to understand the concept of LSC for a common translator like myself. My problem is that the explanations are technical to a degree that they constitute very little help. It's like getting an error message that leaves you clueless as to what you can do about the error. In the case of MemoQ it means that you have to invest quite a bit of time and effort to study the inner workings of it, which I resent doing. Make the program intuitive (adapted to the user and not vice versa) and you don't need elaborate schooling.
The whole issue lies in the inaccessibility of programs because they have not been adapted to the user. The flaw principally lies with the maker, not with the user, is my standpoint.
Even so, thanks for the assistance. I do appreciate it.
Subject:
Comment:
The contents of this post will automatically be included in the ticket generated. Please add any additional comments or explanation (optional)
Chun-yi Chen United States Local time: 12:00 Member (2006) English to Chinese
TOPIC STARTER
I haven't figured out the benefits of LSC
Dec 25, 2011
I don't know if it has anything to do with my language combination (English>Chinese), but it seems when I leave the LSC feature on as in the default settings, I get a lot of "hits" in yellow that only contain fragmented source text but have no target suggestions.
I have turned off this feature for now and it does make the translation results window cleaner. Thank you, Dominique, for pointing us to the right place to adjust the settings.
Chun-yi
Dominique Pivard wrote:
I don't think you need to be a programmer in order to understand how LSC works, especially if you're familiar with other CAT tools: it's just a concordance search performed automatically for you, showing everything longer (in number of words and characters) than what you have specified. The default thresholds (3 words, 16 characters) probably work well with most languages and TM's of normal size.
Subject:
Comment:
The contents of this post will automatically be included in the ticket generated. Please add any additional comments or explanation (optional)
Chun-yi Chen wrote:
I don't know if it has anything to do with my language combination (English>Chinese), but it seems when I leave the LSC feature on as in the default settings, I get a lot of "hits" in yellow that only contain fragmented source text but have no target suggestions.
memoQ has no linguistic knowledge of the target language and therefore cannot tell you what the target subsegment is. Right-click the orange hit, select 'Concordance' and use your knowledge of Chinese to decide what the equivalent of the source subsegment is.
Chun-yi Chen wrote:
I have turned off this feature for now and it does make the translation results window cleaner.
Increase the number of words and/or characters so as to reduce the "noise" and obtain meaningful LSC hits.
Subject:
Comment:
The contents of this post will automatically be included in the ticket generated. Please add any additional comments or explanation (optional)
Chun-yi Chen United States Local time: 12:00 Member (2006) English to Chinese
TOPIC STARTER
Thank you, Dominique, for the valuable information
Dec 27, 2011
I tried with various number of word combinations but still receive empty hits. I do think if a feature is there in MemoQ, there must be a reason for it. I will keep on trying until I start getting "meaningful" hits. In the meantime, thank you so much for pointing me to the right place!
Chun-yi
Dominique Pivard wrote:
Increase the number of words and/or characters so as to reduce the "noise" and obtain meaningful LSC hits.
Subject:
Comment:
The contents of this post will automatically be included in the ticket generated. Please add any additional comments or explanation (optional)
Tom Fennell United States Local time: 14:00 Russian to English + ...
Long List? Turn off the Euroterm plugin!
Mar 2
I am an "expert" (3 day) MemoQ user. The Euroterm offerings were driving me buggy. (I thing there should be an entirely separate window for this, or they should be at the end).
I just turned them off.
Tools>Options>Terminology Plugins
uncheck
Done.
As for MemoQ vs SDL, I'm not sure MemoQ handles inflection as well (and for me, translating from Russian, this is a very important issue), but many of SDL's advantages in regard to auto-entry have no significance: I use voice recognition.
SDL's lack of support for standard voice recognition protocols means seriously sub-standard and nerve-wracking performance, while dictation into MemoQ is absolutely optimal, as far as I can tell after several thousand worlds.
Subject:
Comment:
The contents of this post will automatically be included in the ticket generated. Please add any additional comments or explanation (optional)
Steven Capsuto United States Local time: 15:00 Member (2004) Spanish to English + ...
Agree
Mar 2
When I switched from Trados to memoQ last year, the biggest thing I missed was fuzzy matches of termbase entries and... more importantly... fuzzy matches in concordance searches. Trados did that MUCH better, though on the whole I prefer memoQ.
[Edited at 2012-03-02 06:25 GMT]
Subject:
Comment:
The contents of this post will automatically be included in the ticket generated. Please add any additional comments or explanation (optional)
Start and finish your translations faster than ever with Fluency Translation Suite 2011. TMs, Terminology, and Online Resources are all fully integrated and only a click away. Download a free trial today!
memoQ translator pro is the premium product for professionals. It is Kilgray's best-selling tool among freelance translators: you get all the functionality available in memoQ in your local environment plus the ability to work on remote servers.
The translation workplace 
