Some of the phrases will be translated from a limited set of strings and then reused several times. The phrases have a category-like behavior. Those strings are very well suited for machine assisted translation, that is by using translation memory.
It should be checked if there are work done by Translatewiki that can be reused.
Version: unspecified
Severity: normal
This bug has "need-volunteer" keyword, but it's not clear at all what's being asked.
(In reply to comment #0)
Some of the phrases will be translated from a limited set of strings and then
reused several times.
What are phrases and strings in this context?
It should be checked if there are work done by translatewiki.net that can be
reused.
The Translate extension has translation memory, but translatewiki.net doesn't translate content.
I don't understand what's being translated on Wikidata (most "translation" in it is just abuse of the different article titles for topics in the various Wikipedias), but having two different translation systems (in addition the sitelinks) doesn't seem a good idea. What's called "translation" here may actually not be translation, though... please clarify.
I am guessing what jeblad means here, but my reading is that this concerns the descriptions: if an item like http://www.wikidata.org/wiki/Q328371 has a description in English like "commune in France", and someone translates it to say Croatian, then another item that has the same description in English, e.g. http://www.wikidata.org/wiki/Q269874, could also suggest the same translation for the description in Croatian.
This feature request could be implemented within Wikidata, or via a bot, or via an external site, or whatever. It would help to increase the coverage of descriptions for items in Wikidata.
(In reply to comment #6)
This is a task for a bot.
Let me note that "translation memory" is not a task for a bot, it's provided e.g. by the Translate extension. You're right however that it wasn't clear in the first place whether that was actually needed here.