Wikidata talk:Requests for comment/Automatic labelling

Structure of this proposal

It looks as if things are going to be voted on language by language basis. But I suspect that if we reach different conclusions for Spanish and Swedish it is more likely because the (few) Spanish respondents do not have the same views as the (few) Swedish respondents, than because things really should be different in Spanish and Swedish. I think it would be more productive to focus on the potential benefits / downsides in a general way, rather than directly discussing thing language by language. --Zolo (talk) 20:59, 26 May 2013 (UTC)

I try to interpret the opinions of the community of sv.wikipedia, it is not necessarily my opinion. The community of sv.wikipedia is generally sceptical to that a bot would be able to add a correct label. The rage of the sv.wikipedia-community when they find botcreated wrong labels in 1 of 100 items, is more fatal to their trust to Wikidata, than the benefit of having correct label in 99 of 100 items.

The problem with "no label" in Swedish, can be solved by using fallback-language in the templates on Wikipedia, together maintanance-categories. Adding possibly wrong labels with a bot would remove that option. -- Lavallen (block) 07:57, 27 May 2013 (UTC)

I do not have yet any precise opinion on what should be done, but my point was that this kind of problem does not seem to be language-specific so that it would make sense to have a general discussion rather than fragmenting it by language. --Zolo (talk) 08:09, 27 May 2013 (UTC)

I would say that it is community-specific. If Wikidata starts to add wrong data to sv.wikipedia it would be devastating for the relation between these projects. I am not sure if the same thing is true for es.wikipedia or en.wikipedia. -- Lavallen (block) 08:51, 27 May 2013 (UTC)