Wikipedia

[ciphergoth]

What do you think of Wikipedia?

If you were in charge of a project to create something like Wikipedia, but better, what would you create?

Comments

[ergotia.livejournal.com]

I think it is great, but what I find most annoying is the constant fighting between "experts" over articles etc. However I am not sure what I would do about this - if you assert editorial control over who gets to complete the definitive article you end up with just another encyclopaedia - I guess I like the amateurish "people's encyclopaedia" feel of wikipedia and it may be part of the whole package.

[babysimon]

What I would like to see is something where each word is coloured on a spectrum according to:

* How recently it changed
* How often it's changed recently
* How much the person who edited it can be trusted

Hopefully you'd end up with a version of Wikipedia which visually highlighted contentious sections.

[babysimon]

Oh, and in terms of what I think of it: I like Wikipedia. I think it's good.

[lovingboth]

If this were optional, it'd be very useful.

I'd like to have something like "OK, you've edited this article, you should also have a look here, here and here on the same subject." Some are rather inconsistent.

[faerierhona.livejournal.com]

it's a handy tool, but in places can be extremely inaccurate (I keep seeing lecturer's pissed off at it being quoted as a source of fact) and the arguments annoy me.

Worst though is some of the attitude of people involved! Yes, it's a great tool, but you'd have thought they invented the wheel the way some of them talk!

[keirf.livejournal.com]

I like it. It is what it is.

[xquiq.livejournal.com]

As something of a 'free rider' as far as wikipedia is concerned, I like wikipedia. It's a fantastic way of quickly finding basic information on almost anything. If I need to be 100% certain of the information, I would of course verify it elsewhere, but it's an excellent starting point.

I do however understand the frustrations of those who contribute. There is an inherent problem in that without a high degree of knowledge of a subject it can be a challenge to identify the subject matter expert without the debate degenerating into 'he who shouts loudest'.

I think that improving the accuracy while maintaining the best aspects of the wikipedia approach would be a challenge. I can't see an obvious way to do it without creating greater editorial overhead and thus higher costs and I'm not sure about the idea of selling advertising or similar to pay for it.

In other words, I'm not sure I would create something like wikipedia, because I'm not sure that major changes to the way content is altered or managed could be carried out without losing some essential part of what attracts people to wikipedia in the first place.

[skibbley.livejournal.com]

I don't use it much. I seem to be able to do a quick scan to see if a page looks useful or not and then read further elsewhere to check.

I don't think having a single page for each subject is scalable and sustainable. I'd prefer to have some kind of reputation infrastructure so I can discover the most reliable information using my own criteria or using the criteria of others I trust. One of the things computers are good for is being able to structure the same base data in many ways.

[midwich-cuckoo.livejournal.com]

Wikipedia is in some ways brilliant and in other ways awful

Its difficult to not be impressed with the quantity and depth of content ont he site.

BUT it is very much the world according to reasonably wealthy middle class white folk (although arguably this is no different from Encyclopedia Britanica!)
And it can be a bit like asking your mates down the pub about the question thats been botherting you for a while - just with the difference being that it has more mates than you do.

All in all - a good service - as long as you accept it in its "warts and all entirety" , and dont quote it as a reliable source (but then - since when was anything on the internet a reliable source!)

[drdoug.livejournal.com]

I think Wikipedia is very good at what it does, and turn to it regularly as a first stop for getting an overview of something I want to find out about. Since I really, really like to find out about stuff, I do this a lot. I find the arguments and fighting actually helpful, since it flags up what's contentious among Wikipedians (a proxy measure for other people) - on hotly contended topics I've found the Talk pages an excellent primer on the basic arguments. My life is vastly improved for the existence of Wikipedia. If I was to be marooned on a desert island, Wikipedia would be on the shortlist for my luxury.

Wikipedia is, in my mind, a classic web story: it's rubbish in lots of ways, and runs counter to the ways we know things need to work to work well, but it's nonetheless hugely successful. (Very like, for instance, a certain instance of distributed hypertext.)


If I were in charge of a project to create something like Wikipedia, but better, I would create a way of stopping the project, or folding it in to Wikipedia. Not because Wikipedia is perfect: it isn't. I just don't think "build a better, separate Wikipedia" is possible from here, and even if it were, I don't think it'd be worth the effort.

Actually, the first thing I'd do is go straight back to whoever put me in charge and ask "Better how, precisely?".

[drdoug.livejournal.com]

... and it seems Clay Shirky agrees with me:
http://many.corante.com/archives/2006/09/18/larry_sanger_citizendium_and_the_problem_of_expertise.php
(And also makes some useful points that hadn't occurred to me, and fills in some history I'd forgotten to bring to bear. Again. Damn his eyes.)

[barking-watcher.livejournal.com]

I use it as a quick reference tool, though I don't always to the information as gospel.

If I were to make any changes to an improved version, it would be to try and limit links within articles. I've only seen a few examples and I can't give you any specifics but there have been times when I've seen articles like the example below

Mr Brown was a respected member of the community and lived in a {link} house {/link} on the High Street.

Now that was an extreme example of what I mean but I hope you get the gist of what I mean.

[lizw.livejournal.com]

Yes, that annoys me too. A link to a photo of the house would be great, but if I wanted a definition of "house", I'd use a dictionary.

[reddragdiva]

I will occasionally pepper stuff with links when I picture it being read by the excessively bright ten-year-old encyclopedia devourer I once was.

[thekumquat.livejournal.com]

Given that Wikipedia already exists, all I can think of is a Metapedia - something that trawls published sources for salient facts and notes where they all agree or where there is disagreement, with links to the original sources.

Failing that, a way of distinguishing fact from opinion (anything answering the question 'why' and not related to a hard science counts as opinion), so you could have the 'Metro' version of "$event happened on $date and Xpeople died. $quote1, $quote2", and the extended article which suggests reasons and would be the contentious part ('Evening Standard' version)

[hairyears.livejournal.com]

Yes, it probably would've been nice to think through a system of trust metrics before starting Wikipedia.

But large projects require idiotic optimism in order to begin: and this gets designed-in long before the sensible stuff.

[calum]

I like wikipedia. Its useful, as long as you fact-check it.

What I'd do different, is establish a web-of-trust scheme for authors, so you can see the relative "trust level" of any particular section of text.

Simply marking section of text that are contentious would be good to, as you can always refer to the discussion page to see why they are like that.

[barakta]

I actually like wikipedia a great deal, mainly as a readable (most of the time) thing for randomness - where one just follows links between pages and can hop from subject to subject...

I wouldn't use it as an authoritative source for anything, but as a first-line of information it is useful to get a handle on something quickly. Some wikipedia pages have useful links on them, which are pointing at authoritative and more in-depth sources of information.

I definitely agree it would be nice if there was a system of marking contentious information contained in wikipedia articles, and perhaps a scale of contentiousness vs acceptedness. So the reader could see that the article was in flux (see discussion/talk pages) or something which doesn't really have contentiousness...

I find wikipedia useful to point complete newbies to a subject at, so I have pointed random bloggers at geeky computer things (such as traceroute and what it does and why, which was written from a point of not much knowledge) and for my mum who is a primary school teacher who does 'projects' or has to swot up and provide teaching materials on a completely new subject with very little time to do research.

[dreamsewing.livejournal.com]

I'm going to sound like a Behind-the-times-fuddy-duddy; but I do not use Wikipedia.
I have looked at it a few times when someone has specificaly pointed towards it, that's about it. When I am looking for information I just use the slow road and Google search, looking for information which seems to be as close to 1st person, origin, or trusted ( for example when looking up statistics on health I check WHO).

It's a silly bias based on the first couple of times I looked something up there finding incorrect information which I recognised as such off the bat.

Problems of Wikipedia

[http://users.livejournal.com/_lj_sucks_/]

Problem 1: There's no such thing as NPOV, if you have finite space. Every article inherently selects which pieces of information are relevant and which are not, and that in itself represents a point of view.

For example, the article about typographer Eric Gill pointedly refrains from mentioning his enthusiastic endorsement of bestiality, while the article about Neal Horsley has a whole section on his admission of sex with farmyard animals. While the article about Horsley is under NPOV investigation, it's not the mule-humping that people including Horsley himself are debating the appropriateness of.

To pick another example, should the Minnie Driver article mention that her father was at the center of a $56m insurance fraud involving cash being squirreled away into a Swiss bank account? Maybe it's irrelevant, or maybe it gives insight into why she grew up living in Barbados?

Problem 2: Truth is not a matter of majority vote. In fact, in many important areas truth is extremely unpopular, and the majority of people prefer that it be replaced with bullshit. The sections on evolutionary biology spring to mind, as does the article on Socialist Libertarianism.

Problem 3: Sometimes the most "neutral" point of view isn't nice enough for Wikipedia. As one Slashdot poster parodied, if Hitler was still alive we'd see Wikipedia saying:

"Adolf Hitler was the führer of Germany, who reformed the German economy in the 1930s. He enjoyed painting and playing with his dog. He married his lifelong sweetheart, Eva Braun, two days prior to his death. See also: Criticism of Adolf Hitler".

I'm kinda neutral on Wikipedia. I use it, but I've come to realize that it's a lot less interesting than I thought it was, because of the rule against research. These days it's less about experts writing informative articles, and more about people summarizing information that's already available somewhere else, usually on the web. I'm starting to think of it as more of a specialized search engine with accuracy problems than anything else.

[ladycat.livejournal.com]

I think it's a way for rich patriarchs to share their biased views with the world.

If it was me I'd create a 'thoughtapedia' which explained a topic clearly then allowed you to compare differing viewpoints on it. I'd also make sure it could be inputted to by people without regular access to computers, but I've no idea how.

[ladycat.livejournal.com]

I'd also ask for a bit of background about the person writing the article - are they:

a) Jane/Joe Blogs
b) Someone who read about this in a newspaper (and which one!)
c) A jouranlist who write about this in a newspaper (and which one!)
d) A politician with a viewpoint to push
e) A researcher, and what viewpoint are they coming from
f) An academic, if so where from

etc

[reddragdiva]

http://wikinfo.org/ runs on Sympathetic Point Of View instead, with sets of articles.

Many of the smaller-language Wikipedias are in fact written like this, with people being sought out to write articles on paper if need be.

[reddragdiva]

"I think it's a way for rich patriarchs to share their biased views with the world."

More importantly: what would you suggest, if anything, to remedy this problem? Let's assume for this discussion that Wikipedia isn't going to go away or drop in popularity any time soon, even if that were to be considered desirable.

Also, see this - is this something to address the problems you see?

[reddragdiva]

Thank you for this. Most useful.

[pavlos.livejournal.com]

I would want it hosted outside of the United States for the same sort of reasons that you'd want a version of GPS not controlled by the US or an operating system not controlled by Microsoft. In this case some form of de-centralised hosting may be feasible, but would put extra strain on the means used to synchronise and arbitrate content.

I'd like some sort of arbitration (as in editorial control) of the content. I think in a previous post you mentioned a trust metric, and someone also mentioned a recent change metric, and I like those ideas. However I'm not necesarily after a scheme that decides the authoritative content. For example it may be preferable to expose several versions of an article on, say Cuba: An authoritative factual one, a critical one, a sympathetic one, a generic travel guide, an opinionated travel guide by a foodie/culture/music person, another opinionated travel guide by a sun & sex person, etc.

I'd like to get rid, or at least demote, the piles of trivial content about computers, game characters, and parts of current affairs. I'm not sure how, but the best I can quickly think of is some form of systematic categorisation, so that as a user you can exclude them.

One thing I wouldn't change is the Wiki. It rocks, or at least it rocks harder than any obvious competitors. I'm no expert in this, it's just the one Wiki in comon use that seems free of smileys, sloppy formatting, etc.