Pages in topic:   < [1 2 3 4 5] >
Trados Studio 2017 First impressions
Thread poster: Arianne Farah
Wojciech_ (X)
Wojciech_ (X)
Poland
Local time: 17:08
English to Polish
+ ...
TM view Nov 17, 2016

Dan Lucas wrote:

Walter Blaser wrote:
These features will not work just like this "out of the box". I suggest you read the corresponding instructions on how to do this.

Thank you for that pointer Walter. I checked out the SDL Trados Studio 2017 Migration Guide but this seems to focus on users migrating from old versions, such as 2007, rather than 2015. I can't see anything related to upgrading Studio 2015 TMs.

Regards
Dan


You need to go to the TM view, then select "Upgrade TMs" from the ribbon. Then choose TMs you want to upgrade, click Next, once again Next, and then in the Settings tab tick "Support fragment alignment". Then click finish.

That's not the end. Once the new TM is created, you must go to its settings and in Fragment Alignment option, click "Build Translation Model" (only possible with TMs of 1000 or more TUs).
When the process finishes, click "Align Translation Units" and wait until it's completed.

Do the same thing with every upgraded TM.

Hope it helps.


 
Jesse Good
Jesse Good
Japan
Local time: 00:08
Japanese to English
Are you sure? Nov 17, 2016

Wojciech Matyszkiewicz wrote:

Michael Joseph Wdowiak Beijer wrote:

Wojciech Matyszkiewicz wrote:

Correct me if I'm wrong, but from the exchange with Mr Flanagan, I got the impression that Lift/upLift was supposed to be based on linguistic approach.
However, now I can see that the solution uses statistical methods, which seems a bit disappointing, because you need a fairly large database to receive good results.
Could anyone comment on this? Has this been changed in the course of development?


There was some confusion there. Can't find the ref, but (from what I remember reading in the beta group) basically, the Lift/upLIFT stuff based on linguistic knowledge didn't make it into Studio 2017. It is currently only available/planned for some kind of server thingee. Can't remember the details. Will see if I can find the ref.

Michael


So it's basically a dynamic Autosuggest dictionary with the advantage that you can see the results, which helps in prompting the autosuggest function. A bit of a let down, really.


I believe upLIFT uses the same technology that is in Lift from Kevin Flanagan, otherwise they wouldn't name it upLIFT! Also, statistical translation methods require a much larger amount of TUs then upLIFT. upLIFT only requires a minimum of 1000 TUS (5000 TUs are recommended) that is used to enable sub-segment recall. Probability is *used* in Lift for alignment purposes, but that is not all:

During translation, Lift uses a longest-common-substring algorithm coupled with indexing techniques and configurable parameters (such as minimum fragment length and proportion of ‘stop’words) to match fragments of a query (that is, a sentence to translate) with fragments of TM content, and uses the alignment information to recall and propose the translations of those fragments to the translator.


I believe this is exactly what is in Trados 2017 as it even has the configurable parameters talked about above:



Reference:
http://www.kftrans.co.uk/lift/FillingInTheGaps.pdf

[Edited at 2016-11-17 06:42 GMT]

[Edited at 2016-11-17 06:42 GMT]

[Edited at 2016-11-17 06:44 GMT]

[Edited at 2016-11-17 06:50 GMT]


 
Piotr Bienkowski
Piotr Bienkowski  Identity Verified
Poland
Local time: 17:08
English to Polish
+ ...
Still getting up to speed Nov 17, 2016

1. Installation - smooth

2. Plugins did not migrate despite using the migration tool

3. My largest TM (some 480000 units) failed to upgrade to the 2017 features on one computer. (Some I/O error) The other computer is working on it, and in terms of file size it has a chance to exceed 1 GB

4. Manage to upgrade one smaller TM, but did not notice much in terms of uplift (don't know, maybe units were too short). Autosuggest from fragment matches brings to Studio
... See more
1. Installation - smooth

2. Plugins did not migrate despite using the migration tool

3. My largest TM (some 480000 units) failed to upgrade to the 2017 features on one computer. (Some I/O error) The other computer is working on it, and in terms of file size it has a chance to exceed 1 GB

4. Manage to upgrade one smaller TM, but did not notice much in terms of uplift (don't know, maybe units were too short). Autosuggest from fragment matches brings to Studio what Cafetran already had.

I appreciate being able to fix wrong segmentation by merging across paragraph breaks.

Will keep testing.

Piotr

[Edited at 2016-11-17 07:40 GMT]

P.S. TM upgrade process running since 7 a.m., TM grew from ca. 600 GB to 1.07 GB, process still running...

[Edited at 2016-11-17 09:05 GMT]

P.S.2. Nearly midday local time, upgrade process still running, TM size 1,25 GB. I think I'll have to say goodbye to the jumbo TM era, and treat this file as read-only when it's done.

[Edited at 2016-11-17 10:52 GMT]

PS.3. Upgrade completed. But.... my computer is too slow for a TM of 480 000 that has tripled in size.

[Edited at 2016-11-17 13:11 GMT]
Collapse


 
Oliver Pekelharing
Oliver Pekelharing  Identity Verified
Netherlands
Local time: 17:08
Dutch to English
My favourite bug Nov 17, 2016

My favourite bug is still in there: you still can't select and add several autosuggest dictionaries at once. I've been complaining about this for years, but apparently I'm the only one bothered by it!

Otherwise, I like the look of it so far.

Olly


 
Gene Oh (X)
Gene Oh (X)  Identity Verified
New Zealand
Local time: 03:08
English to Korean
MyMemory Plug-in Not Working on SDL Studio 2017 Nov 17, 2016

I have upgraded my SDL Studio 2015 to 2017. Since then, MyMemory Plug-in has disappeared. Is there anyway to use MyMemory in SDL Studio 2017?

 
Wojciech_ (X)
Wojciech_ (X)
Poland
Local time: 17:08
English to Polish
+ ...
Yes Nov 17, 2016

Gene Oh wrote:

I have upgraded my SDL Studio 2015 to 2017. Since then, MyMemory Plug-in has disappeared. Is there anyway to use MyMemory in SDL Studio 2017?


Copy the plugin manually. Can't help you with the appropriate folders to look for, because I'm not at my computer now. But someone here will help for sure


 
Michael Beijer
Michael Beijer  Identity Verified
United Kingdom
Local time: 16:08
Member (2009)
Dutch to English
+ ...
all the info you need (and more) can be found here: Nov 17, 2016

Dan Lucas wrote:

Walter Blaser wrote:
These features will not work just like this "out of the box". I suggest you read the corresponding instructions on how to do this.

Thank you for that pointer Walter. I checked out the SDL Trados Studio 2017 Migration Guide but this seems to focus on users migrating from old versions, such as 2007, rather than 2015. I can't see anything related to upgrading Studio 2015 TMs.

Regards
Dan


https://signsandsymptomsoftranslation.com/2016/11/17/studio-2017-uplift/ (in Emma Goldsmith's great blog post)


 
Wojciech_ (X)
Wojciech_ (X)
Poland
Local time: 17:08
English to Polish
+ ...
MyMemory in 2017 Nov 17, 2016

Gene Oh wrote:

I have upgraded my SDL Studio 2015 to 2017. Since then, MyMemory Plug-in has disappeared. Is there anyway to use MyMemory in SDL Studio 2017?


Copy the contents of:

C:\Users\YourUserName\AppData\Local\SDL\SDL Trados Studio\12\Plugins\Packages

to

C:\Users\YourUserName\AppData\Local\SDL\SDL Trados Studio\14\Plugins\Packages


 
Wojciech_ (X)
Wojciech_ (X)
Poland
Local time: 17:08
English to Polish
+ ...
It's seems like it's based on statistics. Nov 17, 2016

Jesse Good wrote:

Wojciech Matyszkiewicz wrote:

Michael Joseph Wdowiak Beijer wrote:

Wojciech Matyszkiewicz wrote:

Correct me if I'm wrong, but from the exchange with Mr Flanagan, I got the impression that Lift/upLift was supposed to be based on linguistic approach.
However, now I can see that the solution uses statistical methods, which seems a bit disappointing, because you need a fairly large database to receive good results.
Could anyone comment on this? Has this been changed in the course of development?


There was some confusion there. Can't find the ref, but (from what I remember reading in the beta group) basically, the Lift/upLIFT stuff based on linguistic knowledge didn't make it into Studio 2017. It is currently only available/planned for some kind of server thingee. Can't remember the details. Will see if I can find the ref.

Michael


So it's basically a dynamic Autosuggest dictionary with the advantage that you can see the results, which helps in prompting the autosuggest function. A bit of a let down, really.


I believe upLIFT uses the same technology that is in Lift from Kevin Flanagan, otherwise they wouldn't name it upLIFT! Also, statistical translation methods require a much larger amount of TUs then upLIFT. upLIFT only requires a minimum of 1000 TUS (5000 TUs are recommended) that is used to enable sub-segment recall. Probability is *used* in Lift for alignment purposes, but that is not all:

During translation, Lift uses a longest-common-substring algorithm coupled with indexing techniques and configurable parameters (such as minimum fragment length and proportion of ‘stop’words) to match fragments of a query (that is, a sentence to translate) with fragments of TM content, and uses the alignment information to recall and propose the translations of those fragments to the translator.


I believe this is exactly what is in Trados 2017 as it even has the configurable parameters talked about above:



Reference:
http://www.kftrans.co.uk/lift/FillingInTheGaps.pdf

[Edited at 2016-11-17 06:42 GMT]

[Edited at 2016-11-17 06:42 GMT]

[Edited at 2016-11-17 06:44 GMT]

[Edited at 2016-11-17 06:50 GMT]



I think so. Firstly, if it were based on a linguistic approach, I believe it would have to come packed with a number of bilingual dictionaries.

Also, look here:
https://www.dropbox.com/s/p70ps3hu73x8plq/Uplift.jpg?dl=0

As far as I remember, Lift could recall a fragment from a TM with JUST one TU and AFAIK it's not the case here.


 
Emma Goldsmith
Emma Goldsmith  Identity Verified
Spain
Local time: 17:08
Member (2004)
Spanish to English
Statistical yes, but somehow more intuitive too Nov 17, 2016

Wojciech Matyszkiewicz wrote:

As far as I remember, Lift could recall a fragment from a TM with JUST one TU and AFAIK it's not the case here.


Yes, upLIFT in Studio does that. Check out the very first screenshot in my blog post:
https://signsandsymptomsoftranslation.com/2016/11/17/studio-2017-uplift/


 
Wojciech_ (X)
Wojciech_ (X)
Poland
Local time: 17:08
English to Polish
+ ...
OK, now I'm officially confused... Nov 17, 2016

Emma Goldsmith wrote:

Wojciech Matyszkiewicz wrote:

As far as I remember, Lift could recall a fragment from a TM with JUST one TU and AFAIK it's not the case here.


Yes, upLIFT in Studio does that. Check out the very first screenshot in my blog post:
https://signsandsymptomsoftranslation.com/2016/11/17/studio-2017-uplift/



Could upLift creator perhaps comment on how much of the original Lift has made it into Studio 2017?


 
Piotr Bienkowski
Piotr Bienkowski  Identity Verified
Poland
Local time: 17:08
English to Polish
+ ...
Where are these options? Nov 17, 2016

Hi Jesse,

Where in Studio 2017 are the options that you showed in the screenshot? I can't find them.

Piotr


 
Emma Goldsmith
Emma Goldsmith  Identity Verified
Spain
Local time: 17:08
Member (2004)
Spanish to English
Fragment matching settings Nov 17, 2016

Piotr Bienkowski wrote:

Where in Studio 2017 are the options that you showed in the screenshot? I can't find them.



File>Options> Language Pairs>All Language Pairs>TM and Automated Translation>Search

See half way down my blog post: https://signsandsymptomsoftranslation.com/2016/11/17/studio-2017-uplift/

TU fragments in segment fragment matching is one of the few options that is off by default. It might make sense like that, because it won't work until legacy TMs have been upgraded.


 
Piotr Bienkowski
Piotr Bienkowski  Identity Verified
Poland
Local time: 17:08
English to Polish
+ ...
Scattered all over the place Nov 17, 2016

They are. Now, is it better to make these settings before or after upgrading the TM? My jumbo TM has just finished upgrading.

Emma Goldsmith wrote:

Piotr Bienkowski wrote:

Where in Studio 2017 are the options that you showed in the screenshot? I can't find them.



File>Options> Language Pairs>All Language Pairs>TM and Automated Translation>Search

See half way down my blog post: https://signsandsymptomsoftranslation.com/2016/11/17/studio-2017-uplift/

TU fragments in segment fragment matching is one of the few options that is off by default. It might make sense like that, because it won't work until legacy TMs have been upgraded.


 
Emma Goldsmith
Emma Goldsmith  Identity Verified
Spain
Local time: 17:08
Member (2004)
Spanish to English
Changing settings before/after TM upgrading Nov 17, 2016

Piotr Bienkowski wrote:

Now, is it better to make these settings before or after upgrading the TM?


I don't think it matters. Most simply won't work until the TM has been upgraded.

Don't forget that any current projects need to have the new TM in place, with its updated settings.


 
Pages in topic:   < [1 2 3 4 5] >


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


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

Trados Studio 2017 First impressions







Wordfast Pro
Translation Memory Software for Any Platform

Exclusive discount for ProZ.com users! Save over 13% when purchasing Wordfast Pro through ProZ.com. Wordfast is the world's #1 provider of platform-independent Translation Memory software. Consistently ranked the most user-friendly and highest value

Buy now! »
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 »