Slashdot Mirror
Wikipedia Founder Sees Serious Quality Problems
Juha-Matti Laurio writes "The Wikipedia co-founder Jimmy Wales has acknowledged there are real quality problems with the online project. From the article: 'Meanwhile, criticism from outside the Wikipedia camp has been rebuffed with a ferocious blend of irrationality and vigor that's almost unprecedented in our experience: if you thought Apple, Amiga, Mozilla or OS/2 fans were er, ... passionate, you haven't met a wiki-fiddler.'"
I've debated people here and they use wikipedia facts that were wrong as proof they were right. It drove me crazy... he wouldn't take any other source no matter how many, wikipedia was the spoken word. Yikes.
In a perfect world wikipedia would work, but people aren't perfect, and people have agendas... that is why it will never be taken seriously with anyone outside the community.
I know this was intended as a joke, but it might be good for wikipedia.
Lately I'm finding more "missing" articles than problem ones. Topics that should be there but aren't. Maybe they could have some sort of bounty system to get people to write these missing articles. Of course, that would require paid editors to approve the entires before a payment can be made.
Of course there's a lack of quality. Anybody can come in and edit anybody else's work.
Step 1: Create an account
Step 2: Do whatever the hell you want to the whole place
Maybe a level system ought to be put in place. Create enough new entries and then you can edit other users' work. It's not a perfect solution, but it would cut down on some of the nonsense.
e2 | LJ
Wikipedia usually works, in my experience, especially on popular or controversial articles. Just within the last hour, another editor and I had a dispute over whether "dry mouth" is a negative or neutral effect of marijuana. We went back and forth a few times but we eventually agreed to combine that postive and negative effect lists, and now it is all settled. Such compromise is not always possible but it is much of the time and the system usually works.
"It is not how things are in the world that is mystical, but that it exists." -Ludwig Wittgenstein
A solution I liked was to make the publicly-editable entries into an unstable branch, and to promote versions of pages that have been fact-checked and have been agreed to be up to Wikipedia standards into a stable branch. Redirect anonymous viewers to stable pages if available, and mark each version as to which branch it belongs to.
Regarding Wikipedia itself, I find it to be pretty useful as a repository of widely-known information (dates, names etc), very useful on computer-related information, and perhaps not so useful or reliable on other things. But that's still a net positive. Why the hostility?
Could there be a commercial opportunity in forking Wikipedia, and then having an advertising-supported business hire some editors and professionals to verify Wikipedia articles, perhaps in conjunction with other content? Or perhaps having a university fork Wikipedia and then flag which edits have been verified, or edited, by students or professors of the subjects covered by a particular article? Or perhaps introducing a Slashdot-style moderation system (where you can by default, for instance, only see edits which are rated 5*'s or higher?)
http://amishthrasher.blogspot.com/
To make the project work, the project needs to ensure that people who have committment to the project and it's ideas and expertise in the field have some way of at least removing the agendas and making certain that the facts are as correct as possible. I'm not certain that this does not completely invalidate the whole project, however.
What Wikipedia is, however, is what you'd get when you asked everyone what they *believe to be true* based on whatever basis that they tend to trust in. I don't want that to seem like a put down, or a weakness, however. Most people have a firm basis for what they believe to be true, particularly if it comes from their personal, first hand experience. Therefore, Wikipedia will have tons of good information, and it does. However, when it comes to places where people start reaching farther than they can grasp, it starts to break down. And when those people are obnoxious or stridently unaware of their own limitations, you start getting problems.
What that system needs is a filtering system that lets you have the opportunity to screen out the contributions of people who fit a profile that you feel is suspect. The data should all remain in the wiki, but depending on what you, yourself want to see, you should be able to personalize the editing to match what you can accept. If you feel someone is a wacko (it doesn't mean they are), and they make an addition, you may wish to ignore their contributions in certain topic areas and instead accept the article as it exists without their input.
The worst part of this idea is that people who don't want to see what they don't understand, may find themselves hearing the choir singing to them. However, I don't think you can force people to learn things. It has to be their decision. The best you can do is accept their view point in their submission and then let their deeds speak for themselves and have people choose to ignore them. There should be a peer reviewed filter in Wikipedia that doesn't remove non-expert content, but rather, doesn't let the content of experts be overwritten if there is an agenda involved and that view of the article is viewable if you select the appropriate filter.
I do think, however, it would end the wars where the Wiki is compromised by billions of astroturfers and crackpots under different names. Under this system, what they post never gets overwritten, so they have no reason to go covert. At least, they have no motivation to keep up a running list of fake names and constant counter-editting. There will probably always be the people who post the same things under a billion different names to see if they can get to the most "trusted" filters, but if you are careful, they should never be able to sneak on to your lists.
Obviously, this system would have to be automatic, designed well, and probably require a huge amount of storage space to hold everyone's submissions. But I think it would be best suited to the actual aims and spirit of the Wiki, if it could be done.
Amen. As a society we cannot rely on our own knowledge to teach ourselves. I reccomend that we rely on an older, more advanced race of aliens to do our education and encyclopedia-writing for us. This is surely the only path to a brighter future! Notify the United Nations immediately!
The World Wide Web is dying. Soon, we shall have only the Internet.
Whats truly scary is the number of people defending the use of the Wikipedia as a de facto source of information.
Mathematics is made of 50 percent formulas, 50 percent proofs, and 50 percent imagination.
Wikipedia is worthless, from anything other than a triva perspective. Silly me, I once tried to include literature citations in the entry for Julius Caesar, they were promptly deleted and someone re-entered demonstrably false information.
The real challenge is finding the volunteers to fix all the obscure articles. People work on what they find interesting, and if no contributors find a topic interesting, it's not going to get fixed.
The problem is that a lot of the obscure stuff that *is* there is in areas where geek (or rather nerd) types have interests, and it's not always that well-written. In fact, I think this is arguably at the top of the (otherwised unordered) list of problems with Wikipedia:-
(1) The anal-retentive "fact"-adding tendency. Those who'll add obscure/unused abbrevs to a *disambiguation* page. They don't get that some facts are more important than others, or that simply adding information to an article doesn't necessarily make it more helpful. They'll create lots of small stub articles, when they'd be better combined in a single article (placing them in context). If there's one thing I've learned as I get older, it's that leaving stuff out is *hard* but very important. You can't include everything. And you have to order that information well. The self-indulgent factoid geeks don't know or care about this.
(2) Change for change's sake. I'd be interested to see the amount of "churn" that goes on in some articles simply caused by people changing stuff for the sake of it. It's not necessarily a bad thing; it's just pointlessly wasted effort over a minor issue.
(3) *Potential* subversion by those with an agenda, including professionals. I've seen at least one instance of what appeared to be a PR person editing anonymously. This is dangerous, because most zealots with an agenda are transparent; PR and the like are professionals, and more likely to slip under the radar.
(4) Vandalism; annoying, but usually pretty obvious
(5) Lack of citation. This is very rare, and whilst normal encyclopedias don't normally include citations, Wikipedia's credibility would be much enhanced with more of them.
There are probably more, but my brain is full; that's enough to be going on with...
"Slashdot - News and Chat Sites Deviant". (Click "homepage" link above for details).
I've also noticed a trend whereby people will do stealth vandalism, changing one tiny fact or number. This is far more insidious than the harmless dorks who replace an entire article with "Brent Stevens eats babies". This is clearly an effort by people to discredit the very idea of Wikis.
I think it is more than a casual reference. The beauty and danger of Wikipedia is that anyone can update the content. I would contend that this model is better than any encyclopaedia where relatively few people contribute or review content.
There are many specialists on any particular subject and in the Wikipedia model these individuals can update the site to contain relevant and accurate information. Everyone get to peer review the information. It's the long tail for information.
Making it open and accessible actually improves the overall breadth and quality. It is counter intuitive and many people have been unable to grasp this concept with open source projects.
I don't make predictions, and I never will.
Oh, so that's why it's subtitled "the free encyclopedia."
Why not have a rating system? They should make a rating system, so you could add Informative, Incomplete, Biased, etc, and have articles with particularly low ratings flagged for review (do they do something like this already?).
I think they should lock a lot more articles that are known to be complete and accurate. The definition of, say, orange juice hasn't changed all that much in the last 10 years and probably won't in the next 10.
Working these two concepts in together, I think they should have the 'modifiability' of the article be based on how high it's rated. For just a stub, or no article at all, then anyone should be able to modify it. But if the article is long (enough) and complete, then say maybe only a register with many high-rated articles can change it.
I think the main idea here is to promote and protect good content, but I seriously think they should not do anything to restrict an average joe from exlpicitly adding content.
Anyone else there think I'm on the right track?
Partial Credit: The Engineer's Best friend
"Well, the bridge didn't fall all the way down!"
Silly me, I once tried to include literature citations in the entry for Julius Caesar, they were promptly deleted and someone re-entered demonstrably false information.
Yeah, no kidding.
Point 1. The system doesn't favor true information, it favors whoever can be the most obstinate, anal-retentive, vindictive prick. Take this dipshit, for example. Imagine having a flaming, bitchy drag queen editing your stuff. Not to make it better or more correct-- changing/deleting/removing content just because he didn't like edits to other, unrelated articles you'd done.
Point 2. Then you get the tools that label your factually correct additions as "vandalism". They'll delete whole paragraphs just because they consider the article to be "their" article. This is especially prevalent by the older users towards the newer users.
Point 3. Then there's the "vote for deletion" nazis. See Tverbeek, above. Again, as "revenge" for some perceived past slight, these mental giants will put your stuff up for deletion with the rationale that it belongs on uncyclopedia, this is the typical rationale for deleting topics relating to fiction or pop-culture. Why then, do certain "uncyclopedia-quality" articles (i.e. the Klingon dictionary) stick while others don't? See Point 1.
The theory of relativity doesn't work right in Arkansas.
Point 1 & Point 2: Some interesting comments on Digg, mostly to the same effect as the points I was making.
i.e. "on one entry, me and several friends have inside sources (one being the entry) and when we try to correct it, or correct misinformation that has been posted, the sites owner locks it down or chooses the misinformation over what is even know as fact. starting to distrust information found on there due to personal experience."
Tverbeek was a good example, because he's a royal prick, but he's got no shortage of equivalents on Wikipedia.
And my points are reiterated by one of the Wiki admins there, as well (so no, I'm not "trolling", unless you're also accusing Wiki admins of trolling as well):
The majority of edits on large topics are decreasing the quality of those articles. This is because, for most people, the quality of the article as a whole is taking a back seat to the desire everyone seemingly has to have their imprint on articles. This is turning many articles into long lists of disparate trivia instead of naturally-flowing, high-quality encyclopedia articles. Efforts to stem this and make the encyclopedia more encyclopedic are criticized as counter to the spirit of "openness."
His User page is here.
The theory of relativity doesn't work right in Arkansas.
Same here. I once bothered to spend a couple hours writing some new information in a couple articles. I would say I'm pretty much a subject expert on the subject (from first hand experience, studied about it, op in the only related IRC channel, webmaster of one of the biggest related sites and all). I figured I had done great work, and hence had done everyone a service. Most of it was pretty obvious to anyone remotely involved, but the information was really missing and all...
Only to find out some nobody rolled back all the changes a few hours after, asking me to back everything up with references... Like, dude, you want me to backup statements like "the sky is blue" ???
Sorry, but too fucking bad. I doubt I'll even bother visiting them (or referring them, mentionning them or anything) after that. (I've actually taken my wikipedia links off my websites eh)
Wikipedia works best for geeky subjects. Take a look at the articles (well, more like article hierarchies) for Star Trek and World of Warcraft - you won't find a more thorough or more carefully woven source of information anywhere else.
Well, on the one hand including these very specialized articles is very nice. But on the other hand, there are some costs.
The global namespace has a kind of scarcity (see the disambiguation page for Praxis). If the scope of Wikipedia were as narrow as a paper encyclopedia, you could just look up "Lincoln" and immediately get the entry for Abraham Lincoln. Having so many articles makes it more challenging to store and index data, and the disambiguation pages mean that users have to load more pages to get what they want. This increases not only the time users must spend to find the page they want, but also adds to the network load. Of course, there are probably technological ways to improve the 'finding Lincoln' situation (e.g., an "I'm feeling lucky" search that vetoes disambiguation pages).
There is also a problem that articles of marginal interest may have poor quality and rarely be reviewed. Vandalism to such articles may also go uncorrected for a long time.
The biggest personal annoyance I have with Wikipedia is the incredible amount of specialization and detail found in current events articles. The Cindy Sheehan article is the best example of this that I've run across. It is much more like a reference text for specialists than a general encyclopedia article. In a year or two, even those few people that remember who Cindy Sheehan is are not going to care about the day-by-day account of "Bus Tour - Week 2".
Basically, current event articles wind up looking more like a community discussion board. This is not by itself a bad thing, but it is not what Wikipedia aspires to be. If I donate money to Wikipedia, I would prefer it to support useful articles of general interest rather than political discussions that are of interest only to the participants.
The "tortured prose" of this Register article is apparent in their lack of details on how the Bill Gates and Jane Fonda Wikipedia entries are "unreadable crap" (in Jimmy Wales' words). We're merely told this repeatedly, but the Register never backs their argument (or Wales'). Also, one sees another instance of the double-standards which are tolerated for judging Wikipedia and Encyclopedia Britannica.
If "[s]omething that aspires to be a reference work ought to be judged by the quality of the worst entry" then why are we only allowed to judge one encyclopedia—Wikipedia—on that basis? With such a ridiculously high bar, it's easy to hand-pick articles one knows a great deal about and see if the encyclopedia in question measures up.
Which brings me to the next problematic criticism of these encyclopedias: drawing conclusions by weighing too small a sample. I recall that EB's former editor used exactly one entry to conclude that Wikipedia is akin to filth one is likely to find in a public bathroom (or words to that effect). The Register article's critique centers on reviews of two Wikipedia articles—Bill Gates and Jane Fonda's entries. The only way to reach the conclusion that EB has a "handful of errors" (as the Register says) is to do a survey; you can't judge articles you've never read. It seems to me that a proper review of a large encyclopedia would require a far larger sample size than a "handful" of articles in order to justify any reasonable conclusions about quality, no matter what those conclusions were.
Finally, the Register article mentions a few "respon[ses] to criticism" but doesn't actually critique these responses with a proper explanation. Just because one is told something like "this is what my critics will tell you" doesn't mean you have reason to dismiss the criticism. If one is interested in learning what's really going on, one has an obligation to think about the critique and weigh it on its merits. I "welcome the candour" as well the Register does, but I certainly want my candour to come with examples to back up points. When I evaluate EB using the guidelines I'm told to evaluate Wikipedia by, I come up with the conclusion that EB is merely different from, not better than, Wikipedia. And this conclusion I arrive at without giving any credit to Wikipedia for being free (as in the freedom to share and modify) which EB most certainly isn't. So, if I happen to be a victim of EB's "HUAC", I can't do anything to improve EB without going through the gatekeepers that registered their unwillingness to examine the above topics at all.
Digital Citizen
What wikipedia needs to do is have both "stable" and "unstable" branches of wikipedia, like the linux kernel does.
Make searches default to the stable page, with the option to add in the more recent changes by clicking a button.
This has a number of advantages:
Life is too short to proofread.
Different things are "trivia" to different people. From my perspective, the birthdate and biography of someone who lived hundreds of years ago (except for someone historically significant, e.g. Shakespeare or Caesar) is trivia, while a rundown of the features in the latest World of Warcraft patch is not. I imagine the opposite is true for you. My interests are a closer match to Wikipedia than yours, so I'll use that (bearing in mind that it's constantly in motion and checking the Talk and Article History pages as necessary). You have more historical interest, and so a more conventional encyclopedia is probably a better fit for you. It's no shame to Wikipedia that they lack good information in some areas--simply a matter of specialization.
I suspect that this trend will continue. Wikipedia will continue to expand in geek-friendly and pop-culture areas, while articles one would expect to find in Encyclopedia Britannica will be left mostly empty. If you're looking for the title of a Star Trek episode or a comic book supervillain, check Wikipedia; for articles on Ancient Greece, use Encarta. Most teachers don't accept Wikipedia as a bibliographic source anyway, due to the possibility of students editing a Wikipedia article and then quoting themselves authoritatively. I think that as long as people (including Jimmy Wales, the founder) compare Wikipedia to Britannica and expect it to measure up, they'll continue to be disappointed--they're simply different things with different strengths. That's all there is to it.
Information wants to be anthropomorphized!
what I think the writer of this (imho) crappy article didn't seem to get.. it's not a normal encyclopedia. it's a shared media.. to share information.. not to suck it up. :)
if you go to a restaurant on a date.. you pay for it.. this is more like a free shared cook out where you prepare meals for each other for free, ofcourse you're not going to like all of it.. but you can help people with their recipe's and cooking.. you can still bring a date ofcourse
it's just like with opensource software.. it requires interaction, ah different way of thinking.. and that's what makes it a better product in the -end-