Well you can still view the revision history of an item on Wikidata, as you'd expect. I view the information as being more tied to a specific reference than to a specific revision of the item. I don't think the notion of "orphaned" data is as big of a deal in a database as it is in an encyclopedia. We can monitor the creation of new items the same way that new articles are monitored on the encyclopedia. Especially with historical data, it might not be currently included in any sites that we know, but it should still be there for when people want to make historical charts for reports, school projects, etc. The two methods we have under development to improve the situation are ranks and qualifiers. Ranks let you differentiate between multiple claims about a property as to which one is preferred (likely the one with the most reputable reference) and qualifiers are that extra bit of information that let you differentiate multiple claims in a way that is appropriate for the property (perhaps a date for population values). Do you think these methods will be satisfactory for your concerns?
From: TanchocA@mskcc.org
To: wikidata-l@lists.wikimedia.org
Date: Tue, 2 Apr 2013 14:23:13 +0000
Subject: Re: [Wikidata-l] How does wikidata handle topic redirect/merge/split
Hi Danny,
I'm been on the distribution list since the development of wikidata started and I think what everyone has set out to do and accomplished so far is amazing and will have a profound impact just as Wikipedia has.
I've been quietly on the sidelines absorbing some (I have to admit I cant follow all) the intellectual discussions among the participants.
I do have a thought about this issue of "referential integrity" and "orphaned" data that I'd like to share.
Mediawiki has "what links here" to an article, at least for information residing on the same site. It also maintains what a page looks like at a point in time. Since data referenced on a specific edition/revision of an article can now reside outside
of that article, the intent of the information in the article will be lost if it is not tied to the revision of the associated data when that information changes.
One way that this can probably be handled in some future implementation, if not already done, is to also carry within the reference the timestamp of the referenced data as the reference backwards from the data. It will be difficult and cumbersome for
humans to do this but as the link is stored in mediawiki site, code can be added to make the reference. In that process, it can also inform the host of the data, to add it to the "what links here" so there is a backward reference. To prevent spam and other
issues such as performance, only approved sites (such as wikipedia sites) can be added to "what links here".
Feel free to include back the distribution list in your reply if you see merits in this suggestion.
=====================================================================
Please note that this e-mail and any files transmitted from
Memorial Sloan-Kettering Cancer Center may be privileged, confidential,
and protected from disclosure under applicable law. If the reader of
this message is not the intended recipient, or an employee or agent
responsible for delivering this message to the intended recipient,
you are hereby notified that any reading, dissemination, distribution,
copying, or other use of this communication or any of its attachments
is strictly prohibited. If you have received this communication in
error, please notify the sender immediately by replying to this message
and deleting this message, any attachments, and all copies and backups
from your computer.
_______________________________________________
Wikidata-l mailing list
Wikidata-l@lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/wikidata-l