How do you eat an elephant?

How do you eat an elephant?  Well, of course we don’t really want to eat an elephant, we love elephants!  But if we did want to eat something that large we’d do it in bite sized chunks!  This is something that is particularly relevant when working with large resources such as translation memories or terminology that is freely available from a number of places on the internet.  I’ve addressed this before in various articles… these two are particularly relevant to the topic I’m addressing today:

Since writing these articles you can find at least two places where you can get hold of curated terminology pulled from IATE and used in SDL Trados Studio while you translate:

Both of these resources available through the SDL AppStore make it possible to use the massive terminology resources available in any language pair supported by the EU and they have become an invaluable resource for many users.  However, they both suffer from the same problem of not being based on the current state of the IATE resource, and they don’t contain as much of the useful information that is available through IATE itself. The IATE teams are regularly maintaining this resource as you can see from the details they publish on their website:

It’s unrealistic for any of the existing providers of this data to update their downloads on a daily, weekly or even monthly basis, so resolving this can only be resolved by connecting directly to IATE and viewing the current information as it’s maintained on a daily basis by the IATE team.

The IATE API

At the end of 2018 the IATE team made the API publicly available so that any developer could interrogate the data they have available directly without having to download and manually curate huge terminology resources.   This was very exciting news of course and we moved onto this immediately so we could create a terminology provider for SDL Trados Studio.  We built the plugin fairly quickly and then ran a Beta with a few hundred users for several months making sure that the solution was going to be useful for Studio users and also acceptable for the IATE team.

We released a production build of this plugin earlier this month and it’s now freely available on the SDL AppStore for users of the 2017 and 2019 versions of SDL Trados Studio.

The IATE Real-time Terminology plugin

The plugin is installed by double clicking the sdlplugin file you download (a small download!) from the appstore and this will then give you access to several features that are really useful when translating:

  • term recognition in Studio based on the full IATE database
  • autosuggest from IATE
  • view term details from IATE
  • search IATE

Adding the IATE plugin to your project is simple, you do this in the same way you would do it using MultiTerm except you select “IATE Terminology Provider” instead of “File-based MultiTerm Termbase”.

You will then be presented with the ability to filter the results based on the Domain and/or term types.  This isn’t necessary, but if you do find you are receiving a lot of duplication this can help to narrow down the number of results.

Once you’ve done that and you click “Save Settings” you’re good to go!

You can use this provider with as many other termbases, or other custom providers as you see fit.  It’s not one or the other.

Term recognition

Once you’ve added the provider everything will look quite familiar if you’ve used MultiTerm before.  The segments you are working on will show all recognised terms in the source with a red overline and the matching terms will be shown in the “Term Recognition” window.

AutoSuggest

AutoSuggest is enabled automatically if you have it enabled for Termbases under File -> Options -> AutoSuggest -> Termbases.  You will then be able to use the suggestions from IATE interactively as you translate:

View term Details from IATE

This was perhaps the most challenging part of the plugin to handle.  The IATE database contains almost 8 million terms and over a million terminology entries so displaying them all in the termbase viewer wasn’t really an option.  Instead we resolved this by returning results that are relevant to the source segment you are in… essentially like eating an elephant!

You might be thinking “isn’t this the same as the term recognition window?”  It’s certainly similar but it’s not the same at all as you can see in this image below where there is only one term found., “entry requirement”, but many entries in the viewer.

The termbase viewer does several things:

  • displays possibly relevant information based around the terms found in the active source segment
  • groups duplicates that you can expand on in the viewer
  • shows other definitions and metadata

You can access the termbase viewer for a particular term by right-clicking on a term in the term recognition window and select “View term details“.

The layout is deliberately very similar to the old flags layout in MultiTerm so it will look familiar to MultiTerm users… and it’s my preferred layout anyway, nice compact and clear.

What you can’t do

You can’t add terms to this solution.  The termbase is read only as it’s owned and maintained by the IATE team.

If you are using your own termbase as well as this while you work, which is highly likely, then you will find a few restrictions that are due to Studio/MultiTerm lacking functionality many users have wanted for many years:

  • to be able to add terms to your own termbase you have to make it the default.  This is needed because you can only add terms to the default, not to a termbase of choice.  Fixing this would be on my wishlist!
  • to be able to “View term details” the termbase has to be the default.  You can only view details of the default termbase.  This is a huge oversight and again something users have wanted for years… with good reason.  Fixing this would be on my wishlist!

So if you are using your own termbase in addition to IATE you will get term recognition and autosuggest capability, and this is definitely very useful, but you won’t be able to use the handy termbase viewer unless you change the default termbase.

Fortunately we have some very clever and enterprising users, so Nora Diaz created this useful AutoHotkey script that allows you to change the default and add a term to a secondary termbase.  It would probably not be too hard to adapt this script so you could also select the termbase viewer for the termbase of your choice.  Great to have these options but I do wish we would fix it in the product!

Search IATE

The last thing this very useful IATE plugin does is it allows you to search the IATE website directly by right-clicking selected text and providing the results in an internal view in Trados Studio:

The viewer is just rendering the IATE website inside Studio so you don’t have multiple seacrh windows open and you can avoid the distraction of your web browser while you work!

Finally…

We also added some IATE related features to “Tell me” in Studio so once you have installed the plugin you can type IATE into “Tell me” and you’ll get a few useful links including places to get some help or provide feedback on the plugin as you use it

All in all this is a very welcome addition to the SDL AppStore and I believe will be used by many translators working in the EU language pairs.  Thanks to IATE for providing this, and even more so for publishing the API.  Developers rule!

0 thoughts on “How do you eat an elephant?

  1. Hi Paul

    Great article.

    I downloaded the IATE Terminology from the app store, version 2017, as I am still not good friends with 2019. I double click, but there is nothing installed in 2017. I have tried with SDL Trados 2017 closed and open. I have right clicked on the plugin and chosen Run……..can’t seem to figure out what I am doing wrong. Any suggestions?

    Best,

    Bo

  2. Hey Paul,

    just installed the plugin for tests in my 2017 version, and noticed that it is really slow. It seems to get a little bit faster when I limit the categories, but still not something one would expect.

    Is it just me, or a know problem with this plugin?

    Thanks.

    Łukasz

    1. Probably not a problem with the plugin and more one related to your internet connection. You could also try making sure “Lookahead” is enabled as that might help in this case after you work through the first couple of segments.

  3. Hi Paul. Your article was very helpful, and I have installed the IATE plug-in with Studio 2017 and can see the recognition/results windows as you describe. However, the error message “Collection was modified; enumeration operation may not execute” has since popped up 3 times in just 2 hours. I suppose it could be a coincidence, but I’ve never seen this error message before. I found a question/response in the Community Help which suggests this type of error could be related to MultiTerm. And I can also see that ‘IATE’ is mentioned in the error details (I’ve saved the xml file if it would help to send you that?), but I don’t understand the content myself so can’t get any further with resolving the issue. Any pointers would be appreciated!

    1. Hi Hayley, the best approach is to post into the SDl Community where you can add the error xml file as an attachment. The forum renders the xml nicely and it’s easy for us to read and see where the problem might be coming from. We can also engage with you better in there. The best forum would be this one – http://community.sdl.com/appsupport

      1. Ok. Is there anything in the xml file that might reveal any confidential info (source document content for example?

        1. No… but you can review the file yourself and just remove anything you’re concerned about before you share it. The only thing you might be worried about is it does have your computer name and the name of the logged on user at the bottom.

  4. Hi Paul, I’ve tried to install this on two computers (Studio 2019 15.1.3.55678) but have yet to get anywhere. The error message “Unexpected character encountered while parsing value: <. Path ", line 0, position 0" appears.
    Is this a known issue?

  5. Hi Paul,

    thanks for mentioning my IATE Language Pairs in your article. It urges me to highlight some points and differences between the IATE approach by SDL and my approach:

    Termbase only, IATE Terminology Provider App (ITPA) or Termbase and Translation memory, SanTrans Language Pairs (SLP)(SanTrans).

    • First of all, the impossibility to make use of Concordance Search and Fragment matches (or Uplift) is (not only in my view) an important drawback for using ITPA.

    You may argue that, taking in account the zillion number of possible sentences, the chance of finding any of them in IATE is practically zero. However, this is not true for example for those translators, who are working for EU-institutions – see the reviews of Olav and Maria.
    For highlighting the differences, I have selected some short sentences from IATE, with quite general content. I have purposely changed the sentences so they aren’t 100% copies of existing IATE entries, but still can be found as real sentences in the internet. You will find them below.

    • An important recent addition to IATE, at least for translators of chemical, biological, medical and juridical documents is the addition of terms in Latin (40847) and Multiple languages (5779). These are not offered in ITPA, however SLP has them in ALL source- and target entries.

    • Another important drawback of ITPA, at least for some users, is the need for internet access – the app can work only with an online connection to IATE.

    • A minor (maybe for some translators important) difference is the translation of the Subject Field into the Target language: this is offered by SLP, but not by ITPA.

    • Next, the time after which the information in the Term Recognition window is displayed: for SLP it is almost instantaneous, for ITPA it could easily take up to 8 seconds. Of course, you can cut out Domains, but doing that, you might miss some info that might be available but is in another domain.

    • You also mentioned the lack of actuality of SLP, while ITPA is always up to date. On average, with an update frequency of ITPA of 1 new revision per year, this is in the order of 4%. When I will sell more packages now that I am visible in the AppStore, I might consider to make two releases per year available.

    • There is however also a positive note to this: it is possible to use the IATE Terminology Provider App and SanTrans Language Pairs next to each other, so you will enjoy the fruit of both (of course with the time penalty.

    Example sentences:
    capital tied up in inventory
    car with very high emission
    methane emitted into the atmosphere
    adjustment of the steering wheel
    centre for international policy studies
    Associations in supporting the fight against drugs and organised crime
    Cinchona officinalis

    1. Hi Henk, thanks for your explanatory clarification. I agree that using your download converted to a Translation Memory may offer some benefits, but I don’t really want to get into a debate with you over this. I have no idea how many translators routinely convert their terminology into Translation Memories, I suspect very few who use Trados Studio as this would defeat the purpose of a true terminology solution where it’s unlikely a concordance search would actually return anything that was not already found through term recognition or a search of the termbase. I think the main benefit of using a converted download is the ability to work offline. Translators also have the ability to work with the EUs translation memories (DGT and others) and I suspect these may be more useful for sentences than relying on IATE which really shouldn’t have them in there in the first place.
      The main thing is it’s good to see three potential solutions on the appstore now, all of which avoid the user having to deal with the intricacies of handling such a large dataset on their own.

Leave a Reply