Not Considering

AEM Connector - "Update Translation Memory" functionality

Hi all,

this is an idea specifically for the connector between TE and AEM.

AEM has a functionality called "Update Translation memory" which is currently only available for use with the AEM-native machine translation. It allows you to make a small change on translation of a live page, and this change will then be saved directly into the connected Translation Memory.

As it is currently, if we want to make one small change to a page, we need to either send the whole page as a project over to Trados, or make the change TWO times - once on the live page in AEM, and once in our Translation Memory in Trados to retain it for the future.

If the AEM connector was somehow directly connected to our Translation Memories in Trados, it would make these small changes much easier and faster to make.

Screenshot of a content management interface with a navigation sidebar on the left, showing folders such as 'Language Masters' and 'Website'. A highlighted button reads 'Update Translation Memory' with an information icon and a red error indicator.

Screenshot of a 'Translation Memory Update' dialog with two columns labeled 'Source Text' and 'Updated Text'. Rows display source text and corresponding updated translations. A 'Close' button is visible at the top right.

Vera Malysheva

  • Dear Vera, we have been looking into this functionality a few times before. Whn clicking on the update translation memory button, AEM will generate a TMX file that includes the original source content, the updated target content, the name of the translation job, and the path and the property that was updated. While it would be technicall possible to just simply update the TM with this update, our service delivery teams see this as risky and still see it necessary to use a linguist to update the TM to ensure linguistic consistency. RWS recommends that any linguistic changes be made as part of the linguistic process.

    Additionally, since a resubmission of the whole page would essentially achieve the same result but with added benefit of the linguists receiving the full context, we do not think this feature would be beneficial.