Talk:Reliable Wiki and Structured Information

From Meta, a Wikimedia project coordination wiki

Add your comments and discussions below[edit]

From what I can tell, this seems to be multiple proposals:

  1. Attempt to recruit "experts". How this could be done is not very clear.
  2. Implement w:WP:Flagged revisions on Wikipedia.
  3. Make statistics easily accessible in articles. (I don't think we need any extra technical work for that. It's perfectly possible to do that already with bot-updated templates.)
  4. Add a whole lot of complicated technical stuff to make WP look really cool.

That's my take on what this proposal seems to be. I might be over-summarizing a bit, and slightly distorting some of this. If I'm totally mistaken, please point it out to me. --Yair rand 22:17, 26 November 2009 (UTC)[reply]


Thanks for the comments, Yair. Yes, it's a sort of longish proposal with multiple components. I have no issues with considering it as multiple or breaking it up into 2 or more proposals. But these seemed to me most important aspects that Wikipedia needs to be improved on. Hence to keep things together I have included under the same title. Below are the replies to the points raised by you:
1. Well, the best procedures for recruiting "experts" in Wikipedia volunteer base is something that we will learn with time and experience. To start with, following are a few ideas. More ideas from volunteers will be welcome and can be experimented with.
  • Identifying the initial set of experts may be a little tricky and little subjective. People well acclaimed and experienced in certain fields will be ideal "experts". One obvious target will be professors in academia (as I have mentioned in my original proposal). We can also have professionals with a degree in a certain field and some experience as "experts". University web-sites, company web-sites, etc will be ideal places where we can start looking for such "experts".
  • Sending out formal letter/emails of invitation from Wikimedia Foundation to potential experts initially identified will be a good way of trying to recruit them as volunteers. The email will explain the new initiative of Wikipedia to increase reliability of its contents and hence the need of "experts" in the various fields. Assuming we send emails to all qualified professors in all Universities around the globe, even if one out of ten of them get back and express interest in volunteering, that will make a huge volunteer base of experts. But I am sure a lot more than that believe in Wikipedia's philosophy of "free knowledge", and hence will show interest to help.
  • Recruiting through personal requests will also be a good way. As exemplified in my original proposal, the present Wikipedia users/volunteers may request potential experts they personally know (for example students may personally request their mentors, professors). This may potentially have better positive response.
  • Once we have an initial base of expert volunteers, it can expand through references (i.e. "experts" refering to new experts).
2. The concept I have presented is different from w:WP:Flagged revisions in the sense that the flagging is not done after a revision is made. Rather a new revision is held for review before it is approved by experts. This may be similar to the proposed "full flagged protection" in Wikipedia:Flagged_protection. However the key differences in my proposal are:
  • The role of an 'administrator' in validating an edit will be taken up by an 'expert' in the field.
  • A pending edit can be approved by only an expert in that field/topic (experts need to be associated with category, topic, portals for making this possible). Accordingly, the review request will be notified to only those concerned experts.
  • The 'expert' will agree to leave a signature after validating a content so that the article may be reliably referenced in academic works since it has undergone review by an expert.
  • Some additional features like rebuttal, experts discussion page, etc will be there to facilitate the process.
3. Templates, according to my understanding, is not a clean way of achieving what I have proposed. It's more expensive, not easily implementable and not adaptable to big changes. Looking up a central database for data is the cleanest and easiest way of maintaining consistency among multiple pages. If models like templates would have done the work of dynamically generating contents, then, for example, HTML templates would have been sufficient for the whole of internet and no one would have gone for Dynamic Web Pages or CMS (the models of which is very similar to what I have suggested). But the fact is now almost all web-sites on the Internet are Dynamically Generated, including Wikipedia. Accessing contents from a central database and maintaing the huge contents is always a challenge and hence the whole subject of DBMS has been developed. Structured database is more user-friendly and hence preferred for a system like Wikipedia where we should keep things simple so that people can use it easily. That's the reason I have proposed the "cool" structured way of accessing the database items.
4. The future of technology & science is always "cool". But one should be able to see beyond the coolness and appreciate the benefits. I have listed a lot of benefits and reasons behind the proposed initiative. The coolness is just a superficial outer appearance.
Subh83 02:30, 27 November 2009 (UTC)[reply]
I think that making statistics readily available in templates is actually possible, and probably not to difficult. Rather than, as you suggested <<world.country(usa).present_population>>, it would be {{world|country|usa|present population}}, updated regularly by a bot. The mathematical functions you suggested are already possible with the regular parser functions. Most of the "structured information" is possible, just not widely used on Wikipedia. --Yair rand 03:19, 27 November 2009 (UTC)[reply]
Perhaps you should read mw:Help:Extension:ParserFunctions. --Yair rand 03:43, 27 November 2009 (UTC)[reply]