2023-10-25 - Translation User Group Meeting (Atlanta)

1

13:30-13:45 Welcome

@Ole Kristian Våge 

2

13:45-14:15: Reference set with Latin terms  

@Ole Kristian Våge / @Rutt Lindström 

@Rutt Lindström: There is Latin in the international edition. However, it is marked as English, so it is difficult to tell the difference between Latin and English unless it is done manually. Latin terms have been copied and pasted in the Estonian language reference set, where it has been marked as Estonian, which is equally wrong. This also happens in other languages.

The issue is not new, but now there seems to be a community effort to create a community content with Latin for everyone.

@Ole Kristian Våge: Interested parties in creating Latin content: Netherlands, Estonia and Norway. In Norway, Latin is mainly used in body structure hierarchy, while it is also used in the organism hierarchy in some languages.

@stefan.schulz : In German, Latin is used in body structure and conditions (differences between Germany and Austria). Latin is not used in the English body structure hierarchy. Important reference: Terminologia Anatomica.

@François Macary : There is need for Latin in the organism hierarchy (microorganisms) in the Common French translation. In the Common French language reference set we are able to add descriptions from the US English reference set, so the same would apply for Latin. The Common French also rely on Terminologia Anatomica.

@Camilla Wiberg Danielsen : Some Latin in the body structure hierarchy and in organisms. But it differs from specialty to specialty. But what is the use case for marking descriptions in Latin?

@Frank Geier : Important, quite a lot of work for the German and Austrian NRCs. A lot of copy and paste, a lot to maintain, so a Latin language reference set would be useful. Germany would like to join a project group. But what about the hybrid forms?

@Ole Kristian Våge : We leave hybrid forms out. Regarding user case, there is a need for new countries doing translations to access descriptions which are Latin from “a reservoir”/language reference set, and these should therefore be marked as Latin. Another user case would be QA of existing Latin content for countries which already have Latin in their language reference sets.

@stefan.schulz : Do we just consider Latin terms as loan words? Like English terms. There should be a common source, and the criteria should be straight: It is the same word used in several languages. Hybrid words are no longer of interest.

@Ole Kristian Våge : Should we start with a pilot?

@Rutt Lindström : Action points from Manages Service, Snomed International volunteered to start the assessment of how much Latin there is in the international edition.

@Ole Kristian Våge : First look at what Snomed Int provides, and then look at it. We have to create a project group: Netherlands, Estonia, Norway, Germany. The following countries can´t contribute actively, but would be interested: Denmark

@stefan.schulz : One approach is to take the English and Spanish versions of the description tables and filtered out what was identical in the organism hierarchy. These would be loan words such as Latin. Can provide a list if deemed useful.

@Camilla Wiberg Danielsen : How to present it? Should there be preferred terms in Latin, should there be a namespace for itself?

@Ole Kristian Våge : Each country decides how to represent it in their own language reference set. But Latin in a community browser?

@Maria-Cornelia Wermuth : Latin should be the preferred term according to international language guidelines.

@Ole Kristian Våge will reach out to Snomed Int. regarding how to do this

@Ole Kristian Våge  will create a project group and invite interested members.

3

14:15-15:00: Machine translation: Experiences from Germany 

@Frank Geier  

4

15:00-15:30: Break

5

15:30-15:45 Work flow for submitted issues to Snomed International

@Monica Harry 

There is a Translation slot in the Content Request System, so the submitted issues will not count regarding the maximum accoring to each countries Service Licence Agreement.

There is a Content Request System for new concepts, new synonyms and additional content.

Fresh desk (info@snomed.org) is for broader issues (not a specific concepts), such as questions regarding Editorial Guide, modelling questions and general questions regarding translationsThere is a central bank in Snomed Int. with questions. 

All projects are open on Confluence. There is an Early Visibility for future actions, possible for comments/concerns. A Content Tracker with old issues is being reviewed now.

Content Issue Management: Smaller issues are resolved as they arise. Weekly meeting in an editorial group, difficult issues to an editorial panel with senior terminologists. A content tracker is created if an issue can´t be resolved.

Information

6

15:45-16:45: Challenging translations types: Cases for the  e-learning pathway

@Ole Kristian Våge 

Discussion/feedback: Cases for translation (This document might be revised/changed before the meeting).

This document contains several cases of problems that translators (and reviewers) encounter when translating SNOMED CT concepts. Although the cases include possible solutions, the main idea is that these cases should be resolved by the NRC (or corresponding organ) instead of each translator (or reviewer) making the decisions. The cases represent problems such as difficult grouper concepts, concepts where the FSN is not in alignment with Editorial Guide and/or modelling, culture specific concepts, difference in language granularity, and not equivalent synonyms in the international edition.  

Input from group: Look at "neck" example again. False friend.

Ole have a second look at "neck", and will pick ready issues and send to @Kath Priest  for publishing

7

16:45

@stefan.schulz 

The Recommendation is using caret ^ for superscript and greather/less than <> for superscript. Also unicode. Problem: For the current tooling it doesn´t work. Exotic characters are not wanted, according to Editorial Advisory Group. In translation, however, super-/subscript is allowed. -Issue has been raised in the German translation user group.

@Maria-Cornelia Wermuth : Translation guidelines follow Editorial Guide. 

@François Macary  :Common French don´t deal with super-/subscript

@Ole Kristian Våge : Maybe we should revisit the guidelines

@Ole Kristian Våge will put the item on text meetings agenda