First impressions on Language cloud terminology

I have tried language cloud terminology on Trados 2017 SR1 and I would like to share my first impressions.

I have converted one of my Multerm termbases to Excel and then uploaded it to Language Cloud and it worked, but it was not easy to map correctly my term note fields. Uploading a Multiterm termbase (.sdltb file) as it is (without conversions) to language cloud would be much easier (also viceversa).

I had too few terms recognized in the term recognition window of Trados Studio with the default language cloud term recognition settings. I have tried to change the settings but now it looks like I have too many terms recognized.

My term notes are not displayed in the term recognition window of Trados Studio. The same notes are displayed with the standard (non-cloud) multiterm termbase.

I find the possibility to share a  termbase for a specified time very useful.

Stefano

  • Dear Stefano,

    Thank you very much for taking the time to try it and give feedback.

    Natively supporting SDLTB to create a LC Terminology database is on the short-term roadmap, so that indeed you don't need to export/import, so hopefully this will become easier soon.

    It would be great to understand why you didn't find the mapping of the note fields easy, maybe something we can improve on.

    In terms of the recognized terms, this is one of the key areas we are looking for feedback on, as we want to try and find the best setting and reduce the number of options exposed. Maybe you could share what settings you find best for your language over time?

    We are also planning to support showing term notes in the future.

    Thanks,
    Luis

    Best regards,
    Luis Lopes | Principal Product Manager | RWS | (twitter) @Luis___Lopes |

  • Luis,

    that is what I did to convert from my local termbase file to the cloud termbase.

    I created an empty cloud termbase with basic template, added a single entry, filling in source term, target term, an entry level note, a source term note, and a target level note, then exported to Excel. I used the first row of the Excel file to map the contents of the Excel file I exported  from my local termbase with Glossary Converter. I mean that I pasted the contents of the individual columns (source text, target text, entry level note, source term note, target term note) from one Excel file to the other. I saved the Excel file and imported in the cloud termbase. Everything is ok, but it is a little bit complicated. I'm sure there are more efficient methods.

    About term recognition:  Character-based fuzzy search finds too few or no terms, whatever setting I tried to change, while Linguistic fuzzy search finds too many terms, whatever setting I tried to change.

    I also would like to add a question related to Cloud Machine Translation.

    Why only Baseline is available here? In the previous release of Trados it was possible to select Baseline + Adaptive Engine or something like that. 

    Stefano