Concordance search problems
Thread poster: Roberto Silva

Roberto Silva  Identity Verified
United Kingdom
Local time: 10:35
English to Spanish
Jun 3, 2005

I am having problems with the concordance search feature in TW 6.5.439 (but also happens in every other 6.X version).

If I search for a word, let say:

file

but the word is present in the TM only in upper case, I get no results. If I search for the same word in upper case, it finds all occurrences (all upper case, of course). I have set the coincidence level to 30% and even under this setting, it does not find the word. The word is present in several sentences, never isolated (as a one word sentence). However, this should not have any influence in the search: if the word is present in a sentence accounting for a 50% similarity and the coincidence setting is set at 30%, it should find the sentence, regardless of whether it is in upper case or lower case.

Additionally, I do not understand why a search like: *ile, or *ile* returns nothing at all.

Anybody having the same troubles?
Any idea on how to set concordance so that it does not account for case differences (irrelevant under most circunstances when you are trying to find a translation)?

Thanks


Direct link Reply with quote
 

RWSTranslation
Germany
Local time: 11:35
Member (2007)
German to English
+ ...
TM Jun 3, 2005

Hello,

try to reorganize the TM.

Make sure, that the database version of the TM is 6.

With kind regards

Hans


Direct link Reply with quote
 

Roberto Silva  Identity Verified
United Kingdom
Local time: 10:35
English to Spanish
TOPIC STARTER
Concordance search problems, even reorganizing the TM Jun 3, 2005

Hi Hans,

I forgot to mention it. This happens irrespective of whether the TM is reorganized or not. I always reorganize TM as a first step in solving issues.

It must be something else. I have even tried exporting the TM, and creating a new one (not even using the Copy button, but configuring it from scratch). It happens in the new TM as well.

Maybe I need to split this issue in two:

1.- The concordance search does not find terms for which there is a difference in character case(upper case, lower case). This difference can be in the initial character (for instance File and file) or in any other character (for instance fiLe and file). This problem does not happen in other TMs I have, but it is not solved by exporting/importing to a new TM

2.- How can I search using wild characters like typing *ile to search for file and File, and typing *ile* to search for file, files, mile, Miles, etc.

Thanks.


Direct link Reply with quote
 

RWSTranslation
Germany
Local time: 11:35
Member (2007)
German to English
+ ...
Wildcards Jun 3, 2005

Hello,

it is possible to search more specific with the data maintenace function.

The other concordance problem is not really clear for me. Maybe you can give an example with the search string and with the TUs which are not found

Hans


Direct link Reply with quote
 

Austra Muizniece  Identity Verified
Latvia
Local time: 12:35
Member (2004)
English to Latvian
+ ...
Similar problem Jun 4, 2005

My Workbench sometimes doesn't recognise words in a different case, also different fonts - and that even happens when it's a 100% match. I haven't yet found a way to solve this:(

[Edited at 2005-06-04 11:55]


Direct link Reply with quote
 

Roberto Silva  Identity Verified
United Kingdom
Local time: 10:35
English to Spanish
TOPIC STARTER
Concordance search problems Jun 6, 2005

Dear Hans, Dear all,

I have carried out a series of test and it seems to me that there is a huge bug in Trados TW Concordance search. At least, in the versions I have tried (6.x).

I have reduced the TM to only 3 segments, just enough to prove my point. I have stripped every other info from the segments (user info, creation date, etc). I have manually changed the original source and target languages, but you can also change it if you like to do so. It does not have any influence in the results (as it should be anyway, because language selection should not alter concordance search). Using the segments as they are right now, you should create a new TM with source language Spanish Modern Sort and target language English UK. The segments are:


Cuida la Naturaleza
Cuida la Natura


Cuida tu de la salud de la naturaleza
Tingues cura tu de la salut de la natura.


CUIDA LA NATURALEZA
CUIDA LA NATURA.


Now that you have the TM, try searching for Naturaleza. Now try searching for naturaleza. As you might see, the results are different, while they should yield identical results.

As I mentioned, this does not happen with every TM and / or every searched term. Just with some TM and / or some terms, but... not finding some terms it is a big issue in a program like this, where finding a translation is a key task for the translators.

I would say that the issue is not related to the TM, but to the concordance search in every TM. It depends on the terms you search for.

Any ideas from the Trados people...

Regarding the second problem I mention, the wildcard searches, this is not a bug, but a big negative point of the product. I am sure any translator that have worked with Trados will tell you that it is not the same to carry out a concordance search as to go through a File > Maintenance > Search or any other alternative method. Translators need a fast and easy way to find terms, even variations. This has nothing to do with having a separate terminology database. I am talking about searching in an active TM for recently introduced terms or sentences. This should be possible though the Concordance search, but it is not. A big miss for a good product like Trados.

Thanks


Direct link Reply with quote
 

Ralf Lemster  Identity Verified
Germany
Local time: 11:35
English to German
+ ...
Response by Trados Jun 6, 2005

Hi Carlos,
Any ideas from the Trados people...


I discussed this issue with Trados Support. According to them, the behaviour depends on how you have defined your automatic substitution settings. If you don't use the acronym substitution feature (uncheck "Acronyms" in File - Setup - Substitutions) then uppercase and lowercase strings will behave differently and should be recognized more reliably. The TM will need to be reorganised or exported/imported for this change to be applied permanently. As it is impossible to predict each individual scenario, users should test this out carefully for their particular cases and requirements.

Best regards,
Ralf


Direct link Reply with quote
 

Anette Herbert  Identity Verified
Local time: 10:35
English to Swedish
+ ...
Just a tip Jun 6, 2005

but the word is present in the TM only in upper case, I get no results. If I search for the same word in upper case, it finds all occurrences (all upper case, of course).

Although this does not solve the problem as a whole, a little workaround if there are no hits, is to change the search word, by using the quick shortcut shift+F3, into upper/lower case or initial upper case, and see the results that produces.

//Anette


Direct link Reply with quote
 

Roberto Silva  Identity Verified
United Kingdom
Local time: 10:35
English to Spanish
TOPIC STARTER
Concordance search problems Jun 6, 2005

Hi Ralf,

Thanks for your reply.

My current TM settings in Substitutions are:

Checked:

Numbers, Dates, Times, Measurements

Unchecked:

Acronyms, Variables

This are the settings I use in most of my TM, the ones I have used with the sample segments I sent to this forum (the problematic segments) and I believe they are also the standard settings in TW. So it must be something else.

Just in case: I tried checking Acronyms, reorganizing the 3 segments TM, and then searching. No luck. Then I unchecked the Acronyms setting, reorganized the TM and tried again. No luck.

I created a new TM, making sure the Acronym setting is unchecked. Then imported the 3 segments. No luck. I tested all this in a different computer. Same results, so it seems it is not computer related. It is not language dependant either.

Try the following searches, with the Concordance > Minimum match value settings at whatever you want (even the lowest, 30%). Remember that the target word in all cases should be "naturaleza". I bet you will be as puzzled as I am.

Naturaleza
Naturalez
NATURALEZA
NATURALEZ
aturaleza
natural
naturalez

The results are erratic, and give no indication to the user as to what is really searching for and where does it find "differences" among the searched term.


Direct link Reply with quote
 

Ralf Lemster  Identity Verified
Germany
Local time: 11:35
English to German
+ ...
Contact Trados support Jun 6, 2005

Hi again,
I suggest you contact Trados Support.

Best regards,
Ralf


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 »

Concordance search problems

Advanced search







BaccS – Business Accounting Software
Modern desktop project management for freelance translators

BaccS makes it easy for translators to manage their projects, schedule tasks, create invoices, and view highly customizable reports. User-friendly, ProZ.com integration, community-driven development – a few reasons BaccS is trusted by translators!

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 »



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