Pages in topic:   < [1 2 3] >
New online tool: fast and accurate EU legal corpus search with references.
Thread poster: robinp
robinp
robinp
Hungary
TOPIC STARTER
good conversation! Apr 10, 2019

@Michael, Josephine: thank you! Integrating IATE added to the queue.

@expressisverbis: thanks! Added PT to the queue

@Samuel: Great observations! Until better clarified on the UI, let me reply here.

If there are any exact phrase matches, we feature those in the results. If not, we fall back to exact but non-phrase matches (requested words scattered in the source text).

For
... See more
@Michael, Josephine: thank you! Integrating IATE added to the queue.

@expressisverbis: thanks! Added PT to the queue

@Samuel: Great observations! Until better clarified on the UI, let me reply here.

If there are any exact phrase matches, we feature those in the results. If not, we fall back to exact but non-phrase matches (requested words scattered in the source text).

For legal translation, the farther we get from exact, the less useful the example. That's why we don't return approximate matches, for example alternate cases (though you could argue "case" is a relatively safe alternative..).

Yes, we treat the apostrophe as separator. This make search flexible for some languages like French. You are right that English might not need that treatment.

Generally legal terminology research is a slow process. We bring significant speedup (not saying we stop here. Searches conclude in 1-3 seconds usually. But as you saw when searching for "hello", we emphasize returning a result, if any exists, over cutting a longer query short. "hello" only occurs at the end of a long document. Due to how we perform paragraph alignment, those matches are slower to retrieve - for now at least.

About the 8 results: that's an artifical limit now. In the upcoming release we'll add total hit count display and pagination (among others).

What is your main translation usecase? What is one feature you would add/change? I'm putting these responses in stacks and see how they fit our vision.
Collapse


 
expressisverbis
expressisverbis
Portugal
Local time: 10:18
Member (2015)
English to Portuguese
+ ...
Great! Apr 12, 2019

Thank you so much, Robin!

 
robinp
robinp
Hungary
TOPIC STARTER
New feature: linkable search results. Apr 16, 2019

Thank you!

In the mean time there's a new release - the search results can be easily linked to by copying the url from the browser. Can easily point someone to the Juremy results. Enjoy!


 
TOMBE Franklin
TOMBE Franklin  Identity Verified
Local time: 11:18
German to French
+ ...
Just for legal texts? Apr 16, 2019

Great,
It is a cousin of Linguee and Reverso. But I wish to know if it better handles legal texts?


 
Josephine Cassar
Josephine Cassar  Identity Verified
Malta
Local time: 11:18
Member (2012)
English to Maltese
+ ...
@ Robin Apr 26, 2019

Thank you. I find it better than Linguee for EN

 
evt9n
evt9n
Belgium
Dutch to English
+ ...
Bit late to the party, but... May 16, 2019

Would it be possible to add Bulgarian? I know it's quite a niche language but your tool saves so much time from manually querying the eur-lex database.
Thanks either way, super useful!


 
robinp
robinp
Hungary
TOPIC STARTER
Your lucky day May 16, 2019

@evt9n: Thanks for the compliment! No language is too small (as long as it's in the EU.. and English, come whatever. We are preparing for a major update, stay tuned. In the mean time subscribe to our news list so you get to know once it's there!

 
Elin Gjerrud
Elin Gjerrud  Identity Verified
Norway
Local time: 11:18
English to Norwegian
+ ...
norwegian to the list? May 21, 2019

Would it be possible to add Norwegian also? it would be very helpfull!

 
robinp
robinp
Hungary
TOPIC STARTER
released juremy.com update with 24 languages May 24, 2019

Now Juremy.com has all EU languages loaded! There are also premium convenience features in case you want to support further development.

@Elin: Unfortunately the Eur-Lex data source doesn't have Norwegian version of the documents as far as I know.


 
Philippe Locquet
Philippe Locquet  Identity Verified
Portugal
Local time: 10:18
English to French
+ ...
Separate columns for selection / tmx creation May 27, 2019

Hi,

Thanks Robin P, very interesting tool!

Now I found a use that is of interest to me with your system: That's to create specific TMs (in tmx). I think that could save on research time.

However, the parallel view makes selection a pain, it gets both sides. Could you make text appear independently on both sides so that text can be easily selected keeping the selection only on one language side?

I had to open in different windows and then all th
... See more
Hi,

Thanks Robin P, very interesting tool!

Now I found a use that is of interest to me with your system: That's to create specific TMs (in tmx). I think that could save on research time.

However, the parallel view makes selection a pain, it gets both sides. Could you make text appear independently on both sides so that text can be easily selected keeping the selection only on one language side?

I had to open in different windows and then all the tables appear again, which makes text selection more painful.

ALIGNMENT:

For those interested, I copy pasted my text to align in an updated version of Wordfast Aligner.
You find it on wordfast.net
Under Products/Wordfast Aligner
Just specify language code on the top of each column, and paste your text.

On the sample from Juremy I uploaded there I go zero misalignments, everythig was correctly aligned! That's because now this aligner uses a special sauce that kicks out any dubious alignment or segment that would be useless in a text sense (only numbers etc.).

So for me Juremy + WF Aligner looks like a good combo if I want to get a specific TM in a blink.
Just please make selection easier

My bests
Collapse


 
robinp
robinp
Hungary
TOPIC STARTER
How about direct TMX download May 27, 2019

@Fi2: Happy you like it! Using the results to build a translation memory is an interesting idea. Some remarks:

It seems you can paste in multiple columns the way you want: hide the title display, select all rows with all three columns (celex, source, target), hit Ctrl-C. Then paste into Excel or LibreOffice Calc - it appears in multiple columns. Then you can operate on a per-column basis further.

However, if this is really a common use-case, I would rather have Juremy o
... See more
@Fi2: Happy you like it! Using the results to build a translation memory is an interesting idea. Some remarks:

It seems you can paste in multiple columns the way you want: hide the title display, select all rows with all three columns (celex, source, target), hit Ctrl-C. Then paste into Excel or LibreOffice Calc - it appears in multiple columns. Then you can operate on a per-column basis further.

However, if this is really a common use-case, I would rather have Juremy output the TMX files directly.

About alignment: I don't think using an additional aligner buys much, as Juremy's results are independent. If there's occassional misalignment, the best one could do is drop the misaligned entries, but - based solely on the displayed results - the good alignment can't be restored. Juremy Alignment is correct most of the time, though there's room for improvement when the originals are severly misaligned. Note: I plan to add peak-around functionality as quick-fix for off-by-one cases.

If you are serious about the TMX output use case, I will consider adding it as a premium functionality. Thank you for the suggestion!
Collapse


Philippe Locquet
 
Philippe Locquet
Philippe Locquet  Identity Verified
Portugal
Local time: 10:18
English to French
+ ...
Copy paste, tmx feature May 27, 2019

robinp wrote:

@Fi2: Happy you like it! Using the results to build a translation memory is an interesting idea. Some remarks:

It seems you can paste in multiple columns the way you want: hide the title display, select all rows with all three columns (celex, source, target), hit Ctrl-C. Then paste into Excel or LibreOffice Calc - it appears in multiple columns. Then you can operate on a per-column basis further.

If you are serious about the TMX output use case, I will consider adding it as a premium functionality. Thank you for the suggestion!


I'll have to try this but I was contemplating avoiding yet another copy-paste step with spreadsheet.

Regarding alignment capabilities, I have to say I make custom alignments professionally for different use scenarios and I know what is going to be useful to me and what's not, I'm looking for accurate and filtered results allowing for less intensive manual corrections.
Tools performing at that level are rare.

My bests


 
robinp
robinp
Hungary
TOPIC STARTER
Many feature updates Nov 15, 2019

Just wanted to ping the thread about recent updates:

- We now display Eurovoc topic markers along with the results, to put the usage in more context,
- have multi-language titles (formerly was just English),
- most importantly, we tuned search speed so queries conclude within a fraction of a second.

Where are we heading now?
- Based on user feedback, we'll begin to load sector 5 of EUR-Lex as well.
- Make the paging work harded to bring a full pa
... See more
Just wanted to ping the thread about recent updates:

- We now display Eurovoc topic markers along with the results, to put the usage in more context,
- have multi-language titles (formerly was just English),
- most importantly, we tuned search speed so queries conclude within a fraction of a second.

Where are we heading now?
- Based on user feedback, we'll begin to load sector 5 of EUR-Lex as well.
- Make the paging work harded to bring a full page of diverse results. (Now sometimes you get only a handful of results, since identical entries are deduplicated - we need to backfill those).

Robin
Collapse


 
Merab Dekano
Merab Dekano  Identity Verified
Spain
Member (2014)
English to Spanish
+ ...
Nice to have, but not innovative Nov 15, 2019

Similar concept and functionality as Linguee.

It saves you copying the text to a search engine between quotation marks and adding “Eurlex”, then moving back and forth across the EU documents in both languages (yes, there were times when I was prone to this sort of self-inflicted masochism).

At the European Commissions DGT, they are using this kind of tool, which pulls the existing translations (parallel texts) of the already published documents. They don’t give y
... See more
Similar concept and functionality as Linguee.

It saves you copying the text to a search engine between quotation marks and adding “Eurlex”, then moving back and forth across the EU documents in both languages (yes, there were times when I was prone to this sort of self-inflicted masochism).

At the European Commissions DGT, they are using this kind of tool, which pulls the existing translations (parallel texts) of the already published documents. They don’t give you access to it when you work for them externally, though. That’s when masochism kicks in...
Collapse


 
robinp
robinp
Hungary
TOPIC STARTER
Execution first, innovation later. Nov 15, 2019

Merab: indeed, that back and forth with the documents is what made us develop Juremy to start with. If you check EUR-Lex 2-4 times per translating page, you can save 15-30 lookups, saving you at least that many minutes per day (if not more). Over a month that is 10 hours saved - not bad, is it?

We are not on the track to innovate yet - first we will bring you a flawless solid experience over the current use case. Then start pulling tasks from the innovation backlog...
See more
Merab: indeed, that back and forth with the documents is what made us develop Juremy to start with. If you check EUR-Lex 2-4 times per translating page, you can save 15-30 lookups, saving you at least that many minutes per day (if not more). Over a month that is 10 hours saved - not bad, is it?

We are not on the track to innovate yet - first we will bring you a flawless solid experience over the current use case. Then start pulling tasks from the innovation backlog

As for Linguee, it is a nice tool, but who knows how often do they refresh their database? Also it is not geared for the legal terminology research usecase (see their EUR-Lex links point to OJ pdfs instead of the nice metadata display we have). That said, some of our users report success by combining their other product Deepl with Juremy - doing a first pass of automatic translation using Deepl, then going over manually and checking suspicious terms from Juremy.

The DGT tool sounds interesting. I wonder why they don't open it to the public. I also wonder if their internal alignments are better than those published on EUR-Lex. We go to great lengths with Juremy to fix up alignment of many docs (and our challenges are not over yet on this front).
Collapse


 
Pages in topic:   < [1 2 3] >


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


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

New online tool: fast and accurate EU legal corpus search with references.






Anycount & Translation Office 3000
Translation Office 3000

Translation Office 3000 is an advanced accounting tool for freelance translators and small agencies. TO3000 easily and seamlessly integrates with the business life of professional freelance translators.

More info »
Trados Studio 2022 Freelance
The leading translation software used by over 270,000 translators.

Designed with your feedback in mind, Trados Studio 2022 delivers an unrivalled, powerful desktop and cloud solution, empowering you to work in the most efficient and cost-effective way.

More info »