Search this!

December 24, 2006 3:44 PM

Wikiasari search engine. Wikipedia founder plans to offer a new search engine using "the same network of followers" for the process. “Essentially, if you consider one of the basic tasks of a search engine, it is to make a decision: ‘this page is good, this page sucks’,” Mr Wales said. “Computers are notoriously bad at making such judgments, so algorithmic search has to go about it in a roundabout way. But we have a really great method for doing that ourselves,” he added. “We just look at the page. It usually only takes a second to figure out if the page is good, so the key here is building a community of trust that can do that.”
posted by Brian B. (29 comments total)

That's remarkably scant on details as to what he's actually planning and, if I were to be cynical, reads remarkably like IT press release journalism. But it's Christamas, and who can be a cynic? It sounds wonderful.
posted by Luddite at 3:55 PM on December 24, 2006

But it's Christamas, and who can be a cynic?

Me.

I'm sorry, I'm supposed to trust a community to tell me what is relevant and what isn't. What an amazingly stupid idea.

Indeed, in America, if half a million people tell me that's the right answer, I can safely discount it as wrong.
posted by eriko at 4:01 PM on December 24, 2006

Eriko, I suspect he's just using it to tell spam from not spam. Which'll work fine until the search engine optimizers figure out how to game it. Sounds like it's mostly a cat-and-mouse game.
posted by swell at 4:06 PM on December 24, 2006

I'll bet the spammers are already signing up to rank their own pages as good.
posted by scodger at 4:09 PM on December 24, 2006

Yeah, he's trading "Is it spam or not" for "Is this person a spammer or not?". If this search engine ever takes off, expect it to come under much heavier attack than Wikipedia does from spammers and such.

(Come to think of it, though, does Wikipedia have a significant spam problem? I've never seen spammy links in a Wikipedia article, but I'm not in the habit of looking through recent changes pages to see what the evildoers of the world are up to...)
posted by arto at 4:11 PM on December 24, 2006

Metafilter: I'm sorry, I'm supposed to trust a community to tell me what is relevant and what isn't. What an amazingly stupid idea.
posted by scodger at 4:11 PM on December 24, 2006

I'm supposed to trust a community to tell me what is relevantThat's exactly what Wikipedia does. Theoretically, articles are supposed to be about relevant ("notable") facts. It's not always perfect, but then, it's amazingly useful.

I'll bet the spammersWikipedia seems pretty good at policing spammers and black-listing sites that abuse. Unlike Google, the blacklist is transparent and open for anyone to see and submit entries.
posted by stbalbach at 4:12 PM on December 24, 2006

I find that most of the Wikia business endeavors to be overrated and they come across as lacking real business sense. I think they are all going to fail, or be cash sinks for the foreseeable future. There is some halo effect (particular with regards to media coverage) resulting from their weak association with Wikipedia's success, but that only goes so far, it will not actually cause these businesses to succeed.
posted by bhouston at 4:13 PM on December 24, 2006 [1 favorite]

Wikipedia spam blacklist
posted by stbalbach at 4:13 PM on December 24, 2006

arto: There are spam attacks on Wikipedia all the time.

Low intensity attacks can be successful, although high intensity attacks will fail. Spamming obscure articles works a lot better than trying to spam the article on SEO or Microsoft or some other high profile article.

One middle of the road case involved a Middle East academic who secretly spammed dozens of articles in his field over many months with links to his articles, more info here.
posted by bhouston at 4:17 PM on December 24, 2006

I'll get excited when he actually produces anything. Wikipedia is one thing, it would be impossible for computers to write all those entries. But organizing search links? I just don't think doing so would give people the same sense of accomplishment.
posted by delmoi at 4:44 PM on December 24, 2006

Didn't we try this before in DMOZ? I found DMOZ to be, well ... sad and pitiful.
posted by adipocere at 5:58 PM on December 24, 2006

Anyone else remember Spider? (Do I have that name right?) It was a search engine that modified itself through community use. When you clicked a result, it took into account how long you stayed at that site, and how many other sites you linked to from it, and used composite data to modify that sites' ranking the next time that search was performed. It was actually getting pretty good for certain topics, but the critical mass thing was the sticking point.
posted by daisyace at 5:59 PM on December 24, 2006

Yet eriko reads Metafilter. Egads.
posted by fire&wings at 6:00 PM on December 24, 2006

Yet eriko reads Metafilter.

Posts, even. A loser is I.
posted by eriko at 6:50 PM on December 24, 2006

You'll end up finding that all the decent, but controversial, pages never get shown as interest groups take over. Search for abortion or scientology or any of a thousand other topics where people get all ideological and the results will be useless.

Since Wikipedia has been much fucked up the administration there, I don't at all trust this secondary project to be any better. I expect search engine topics fiefdoms as currently exist on WP.
posted by Kickstart70 at 7:13 PM on December 24, 2006

People complain about community searches, but its not like the early days of the Internet, where I could surf every page about a subject, there tends to be more content online now.

Someone has to weed that down, and so far google does a good job, maybe Wikia will do better. But static entries does seem to be a deathbed, thats why people stopped using yahoo directory, dmoz, etc.
posted by IronWolve at 7:33 PM on December 24, 2006

Wouldn't it be fun if the spammers were the first to make a sentient AI, just to beat anti-spamming?
posted by Blazecock Pileon at 9:35 PM on December 24, 2006 [1 favorite]

bhouston is right: this will fail miserably, just like every single other wikimedia project since the original.
posted by blasdelf at 4:30 AM on December 25, 2006

will you be able to vote down search results, like on digg?
posted by kolophon at 5:04 AM on December 25, 2006

If the admins are the same assholes as over at wikipedia, I'm not gonna contribute a damn thing. I stopped contributing at wikipedia after being banned for a day for making a grammatical correction to an obscure article.

Never again you fuckers.
posted by aerotive at 6:04 AM on December 25, 2006

But I like pages that suck.
posted by sidereal at 6:46 AM on December 25, 2006

Only if they promise to go easy with the teeth.
posted by nebulawindphone at 8:51 AM on December 25, 2006

It's expensive to corrupt Wikipedia because it's a text repository. You need to create a plausible article or edit an existing one subtly so that your spam has some chance of surviving. AIs can't really do that (yet). Jimmy Wales described his new rating system as "this page is good, this one sucks". How hard is it to get an AI to do that?
posted by Joe in Australia at 1:23 PM on December 25, 2006

wikiasari? what a friggin dreadful name.
posted by dozo at 7:13 PM on December 25, 2006

Jimmy Wales described his new rating system as "this page is good, this one sucks". How hard is it to get an AI to do that?

I agree, his rant about Google being broken is odd considering that that Wikipedia pages usually come up tops on Google, and those pages are the only one's which are instantly corruptible. Wales' own proposed human rating system has never been extended to Wikipedia, so it's not like they have any honest experience with it (and wikipedia could use a rating system).
posted by Brian B. at 9:40 PM on December 25, 2006

This is very similar to stumbleupon, but search-oriented. I find tons of really interesting and relevant stuff on stumbleupon, but it's very random, so I could see a project like this having potential.
posted by nTeleKy at 10:01 AM on December 26, 2006

If the admins are the same assholes as over at wikipedia, I'm not gonna contribute a damn thing. I stopped contributing at wikipedia after being banned for a day for making a grammatical correction to an obscure article.

Never again you fuckers.
posted by aerotive at 8:04 AM CST on December 25

If you used the username aerotive, it looks like you got blocked for uploading copyrighted images of Jennifer Aniston and for blanking your usertalk page.
posted by goethean at 8:26 AM on December 27, 2006

Actually, I think Aerotive was blocked because Yamla made a mistake. Yamla forgot to remove the copyrighted image when he first warned Aerotive. When the image was still present the next day, Yamla thought Aerotive had reinserted it.

Also, blank one's own talk page is a hardly block worthy. I don't think it's ever been part of official Wikipedia policy and I doubt most active editors have ever heard of it. In any case, that's certainly not the reason Aerotive was blocked.

Aerotive, if you had contested the block, I'm pretty sure Yamla would have apologized and removed the block. Mistakes happen, so it's generally best to assume good faith.

P.S. Admins sit at the bottom of the the Wikipedia hierarchy, just one step above registered users. There are about 1,000 admins and some of them are probably assholes. See WP:ADMIN for recall procedures.
posted by ryanrs at 9:20 PM on December 27, 2006