Hoi, When we have no data at all, we have a 100% failure rate. When we import data from Wikipedia we improve on that. When we copy by hand there is an additional failure rate because people make mistakes.
Given the ever increasing number of external sources that are referenced to from many items (often these references are the only statements), it should be not that hard to create bots that compare "our" data with "their" data. The discrepancies found indicate that their is a reason for attention. When there is no discrepancy it can still be wrong..
My point is obvious; adding data to the best of our abilities is to be preferred over no data at all. Thanks, GerardM
On 12 January 2014 20:39, Bináris wikiposta@gmail.com wrote:
2014/1/12 Gerard Meijssen gerard.meijssen@gmail.com
Hoi, When you are able to identify dates of birth / dates of death from WIkipedias, you can update Wikidata wherever there is no dob dod. When all these statements are made, it will be possible to find more people who died or were born on the same date.
Oh, I see. Currently the bot is not able to correctly recognize births and deaths. Rather, it copies a small environment of the original text in a human readable format so that editors can evaluate it.
If we suppose that in a given wiki all the plenty of infobox templates use the same parameter name for birth date, the bot may be tought to recognize it from infoboxes. Then wi still have the question of reliability.
Bináris
Pywikipedia-l mailing list Pywikipedia-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/pywikipedia-l