Slashdot Mirror
Wikipedia's Search Engine Plan
jasonoik writes "Wikia, the commercial company founded by Wikipedia's Jimmy Wales, reveals plans for a new, editable search engine. They say that the goal of the project is to get 5% of the search market. The service does not yet an official release date. The article also leaves open the possibility that the search results may contain ads, and concludes by listing figures of the web advertisement market." Update: 03/11 17:24 GMT by KD : Wikia and Wikipedia are separate companies.
...which sounded delicious.
"Do No Evil" became "Be as corrupt and evil as possible."
An "editable search engine"? Great, now even MORE of the searches I run will pop up ads for v14GR4 and enhancements for body parts I don't possess, nevermind those linkspam sites that just insert the entire fucking dictionary in metacode.
You searched for: Bill Gates
you got: 400 pictures of penises, vaginas, and one picture of a penis covered in something that looks like it came out of the OTHER opening.
Just imagine what all those malcontents out there with too much time on their hands will do with this! It could be truly amusing.
Not *everything* works best when edited by the hordes.
Wikia is not the "company" behind Wikipedia. The Wikimedia Foundation, which is a non-profit foundation, is what's behind Wikipedia. Wikia is a totally separate for-profit company that is run by Jimbo Wales.
Cyde Weys Musings - Scrutinizing the inscrutable
All those bloggers-for-hire that are starting to find themselves unemployed suddenly have new embeded job opportunities.
Wikia, the company behind wikipedia reveal plans for a new, editable search engine. They say that the goal of the project is to get 5% of the search market.
According to Wikipedia, that goal of 5% will triple in the next six months.
The theory of relativity doesn't work right in Arkansas.
Maybe they're first project should be: make wikipedia's internal search work correctly! It can't even handle the most basic miss-spellings now.
If your serious about this, don't compete with google, instead partner with google and make a wiki.google.com provide google's own search results & ads, but filtered and processed in various ways, which are handled by the wiki.
For example, you want to give only unique sites/hits but this may depend upon the host's url.
The Christian religion has been and still is the principal enemy of moral progress in the world. -- Bertrand Russell
I agree completely - the default wiki search needs major, major work. If they get this search software working and add it to Mediawiki, it'll be a major improvement. As a standalone search engine, however, I don't see the point.
What's the advantage of having user-editable search results? Anyone can submit sites to Google already. I don't know the exact statistics, but I'd imagine that most sites that aren't complete trash end up getting accepted - my site is a jumble of code I put together to learn PHP and MySQL, and looks like something out of 1995. It got included just fine. Therefore, the only difference this search engine would have is the inclusion of Google's rejects.
Then we have the editing. Don't get me wrong - I'm a big fan of Wikipedia and believer in the "everyone can edit" system. Nevertheless, I really don't see how free editing can be useful in a search engine. I remember back when Google was first released, one of the things that made it so special is that none of the results were placed by hand. Other search engines placed higher-paying customers at the top (I have no idea if they still do that - I never use anything but Google these days) and consequently the results tended to have problems. User-editing will likely have an even worse affect, with people putting sites that don't belong on top before those that do.
Yes, there will be the community to catch that, but there's a major difference between an encyclopedia and a search engine. In an encyclopedia, there is a limited number of articles, and each one is about a very specific subject matter. There are an infinite number of search possibilities, and very few of them describe only one thing. For example, I'm a big fan of Heroes. Therefore, I go to the search engine and edit the search for "Mr. Bennet" (one of the characters) to list some sites about him before everything else. Then my evil clone, swd09, comes along. He is a big fan of Pride and Prejudice, and changes my edit to list sites about Mr. Bennet from Pride and Prejudice before everything else. I then change it back, and an edit war begins. In an encyclopedia, it could eventually be settled by virtue of the fact that an article is about one or the other. If someone tries to put information about Mr. Bennet from Pride and Prejudice in the Heroes article, it's clear that they're in the wrong. In a search engine, though, how can anyone say whether Mr. Bennet from Heroes or Mr. Bennet from Pride and Prejudice is more important? There's no way to come to a true consensus. To solve the problem, the administration will have to put its foot down and arbitrarily decide, and we end up with a non-user edited system without the neutrality of an algorithm.
Now I search and my results are:
"Tom is a FAG"
"Bilbo Lives!"
"Search engine optimization: do it the Wiki way"
"In the year 1432, the United States of America was founded by Bill Gates and his horde of windows-operating-system killbots..."
The thing that really rocks about Wikipedia's search is the Disambiguation function. Even Google does not have something like this.
They might not realize it, but they already have 50 percent of the search market. At least 50 percent of the "Intelligentsia" search market.
Fifty percent of the stuff I used to "look up" through a google search - I now get through wikipedia. You just have to be smart enough to know that the info you are looking for is most likely in wikipedia. And it most often is. Especially since wikipedia is so open - they've got articles for tons and tons of things that no mainstream encyclopedia would ever touch. I no longer use "fan sites" or "episode guide companies" for the episode guides of TV Series, they're all in wikipedia, and the layout and presentation is even better.
Wikipedia is already a search engine, because the no original content rule means that it doesn't contain anything that isn't summarized from somewhere else, usually somewhere on the web.
GCHQ Quantum Insert installed. If only our tongues were made of glass, how much more careful we would be when we speak
...at least it would get corrected. ;)
Chu vi parolas Vikipedion?
... I thought they were going to fix Wikipedia's search function. :(
You make a good point, but this is mostly true only in english.
In other languages you get much less from the wikipedia.
Problem is this will require a small band of fanatics to do the editing. Now for the "central/core-cultural" stuff that you might expect in an encyclopedia this model may work but web searches are probably more long tail/niche. Not sure that the editing group could ever be representative. Furthermore the risk of bias on small sample size gets even larger. Some of the bias mightn't even be conscious: e.g. exhibiting a preference for a rigourous page over a "dummies guide" (which might be more popular/widely useful).
Much better would be a behaviour based search engine that inferred when users were un/happy with results- e.g. user doesn't come back for more searches or click more links on existing return.Also even say if a user does a "poor" search firstly & then uses "clearer" terms then engine ought in future suggest the "clearer" terms as alt search or even return some of the results. Indeed even better the engine might "cluster" you with other similar users & retunr more relavant results (e.g. effectively inferring that you prefer rigourous complete guides rather than dummies intros).
This would be simpler & actually rely on the wisdom of masses rather than some central command editors, in fact this type of thinking was behind PageRank.