December 23, 2006

Wikia To Launch Search Engine: Exclusive Screenshot

Michael Arrington

159 comments »

Update 12/25/06: Jimmy Wales says this isn’t a Wikiasari screen shot. So what is it?

The Times reported earlier today that Wikipedia founder Jimmy Wales is planning to launch a new search engine next year, to be called Wikiasari.

He’s clearly aiming for Google. He says:

“Google is very good at many types of search, but in many instances it produces nothing but spam and useless crap. Try searching for the term ‘Tampa hotels’, for example, and you will not get any useful results…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.’ 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. 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.”

The new company will be the third business division of Wikia, the for profit company that Wales founded in 2005 and is now led by CEO Gil Penchina. The other two business units are the main Wikia wiki site itself, and the recently launched OpenServing product.

Wikia has raised over $4 million in capital, including a recent round by Amazon.

Despite the fact that the original article reported that Amazon is involved in the project, Wikia is making it clear on the site that they are not invovled in any way (other than as a shareholder of Wikia).

Wikiasari

A source tells us that the working name for the project was “WikiSearch” until recently. It’s clear that Wikiasari will be focused on quality first, depth second. Search results will include tag based navigation, the top three results will be wikipedia content, and the remaining results are determined by sites wikipedia considers to be “reputable” because they are external reference links from wikipedia pages.

Since all search results will be tied to wikipedia, either directly by linking to wikipedia content or because the sites are linked to from Wikipedia, real people will eventually be determining all search results and rankings within Wikiasari. The search engine will be opensource, and the index will be available under a GFDL. Wikia will operate the master version of the index, but others are free to take it under the terms of the GFDL.

The engine itself will be built on the Nutch and Lucene open source projects.

Wikiasari will be a for profit venture, although the “majority” of proceeds will be donated to Wikipedia.

We obtained the screenshot from a trusted source outside of Wikipedia - we can’t guarantee it’s not a fake but our belief is that this is a genuine working screenshot of the application. Click on it to see full size version.

  • Sphere It

Trackbacks/Pings (Trackback URL)

Comments

seems cool

but whats wrong with wikipedia’s regular search feature?

 

Michael,

seems unfair you cant invest in these companies . .. this one is a winner

 

I’m excited for this. This is great for people who aren’t as experienced with the internet. I cringe every time I watch an inexperienced user instantly clicks on paid Google advertisements and not find anything relative to what they were searching for.

 

Unfair that you can’t invest in these companies?

Call them up if you’ve got 7 figures!

 

Without computer automation, how is this any different from a directory? This is not a scalable solution, and a waste of human effort.

 

What a terrible name. Google - easy to remember, easy to type. Wikiwhat? I’ve forgotten already.

 

Well that’s ironic. The first three results of any google search are wikipedia pages.

 

Agreed the name is difficult to remember but this might actually prove to be a winner since if it just searches wikipedia, chances of spam are less but then again whats wrong with the regular search. Maybe they wanna monetize their search as well.

 

The operative phrase is clearly the motto “Searching Wikipedia and its external links” and there will now be even more of a rush to get a link in Wikipedia. For instance, fastluxurycars.com has done quite well in the screenshot despite being a bit on the thin side, because they got a external link under the Wikipedia Toyota Camry Hybrid article. Comparing a Google Search on “toyota camry” at least hints at how weak the example results really are, but paradoxically how much Google loves Wikipedia. For now at least.

 

The name needs to be changed to “wikify”. Wikiasari will never catch, Wikisearch was closer but drab.

Will put a dent in all other search engines imo.

 

Putting the human factor to search engine results will only make the SERPs worse. Another thing for black hat SEO’s and spammers to play around with.

 

The real answer to search and most likely the next google will be people powered search.

Imagine a search engine where you entered in a search term and someone who is online responded with the results for you.

They could promote their own sites or sites where they are affiliated.

To stop people not being helpful for users, the ’search assistants’ get voted of graded as most helpful.

The more helpful you are the better your rating, the more chance to drive traffic to sites incluing yours.

It’s true that people are the best source of knowledge, machines are good at quantity, not quality.

Goodbye Google

JB

 

Great idea. What about:

1) Everyone rushing the Wikipedia and causing what happened DMOZ to happen to them?

2) If Google drops Wikipedia?

3) When implies their own patent for people-powered search?

It will be interesting.

Warmly,

JP

 

The SERPs results on the screenshot are NOT bringing up any Toyota.com pages.

Toyota’s official site is the #1 listing on Google Yahoo MSN and Ask for that search.

Wikipedia is also on the first page for all four majors.

★ ☆ ★ ☆ ★ ☆ ★ ☆ ★ ☆ ★ ☆ ★ ☆ ★ ☆ ★ ☆ ★ ☆ ★ ☆ ★ ☆ ★ ☆ ★ ☆ ★ ☆

It will be interesting to see how they deal with the Link Popularity controversy, and the Banning of spam sites.

 

“and the remaining results are determined by sites wikipedia considers to be “reputable” because they are external reference links from wikipedia pages.”

It will be interesting to see if this will dilute the value of Wikipedia itself as individuals will have an “incentive” to throw more energy into *gaming* Wikipedia itself (with SPAM links) in order to optimize their respective Wikiasari rankings?

 

if you how to start a clothing line my blog…shows up…so I think google is handling its business! :)

 

I’m glad that someone is truly taking on Google. I wonder if Google will adapt if Wikiasari is a success.

 

Aiming at Google and using its search ads and results - it sounds SO RIDICULOUS. He is BSing his investors :) What about Semantic Web?

 

The names “Google” and “Yahoo” might have been criticized initially, but they’re easy to say and they roll off the tongue. Unfortunately “Wikiasari” just seems like a completely awkward and poor name for a search brand. But at least you can call the users Wikiasarians.

For what it’s worth, I agree with Oliver. Don’t see anything novel here and I’m always amused when a company’s model revolves around phrases like “community of trust.” It’s amazing that some people actually buy into this stuff. If this thing does get traction, it WILL be abused very quickly.

 

Nice idea, but couldn’t they come up with a more original page design. At first glance, it looks just like Google.

And they definitely need to work on a better name. Wikisearch isn’t that bad.

 

@ Comment #14: Actually, they are.

fastluxurycars.com/Hybrid-Cars/2007-Toyota-Camry-Hybrid.htm is linked from en.wikipedia.org/wiki/Toyota_Camry_Hybrid for example.

This seems more like a replacement for the current Wikipedia search than a real competitor to Google, et. al.

 

Scratch that, I misread your comment. :) And in fact, you make a good point. Toyota’s official Camry page should be the first result in a search for Camry, but Wikipedia doesn’t apparently link to the Toyota site (at least not the US site–the main Toyota page on Wikipedia links to the Japanese site). That’s a fatal flaw in the design of this engine.

 

Mmm I just don’t get it…

Since when the wikipedia become a websites directory?

I thought we already had del.icio.us for that… :/

 

Asari is a Japanese word for “rummaging around”, but the word has some slovenly connotations and is used in context with, for example, homeless rummaging around or less than ethical people “searching” for prey… wonder if this will hinder a launch in Japan, a country that prizes their hierarchical vocabulary?…

 
 

So, if this works out, it may be a threat to google. Unless, of course, Google buys Wikia… Googlepedia anyone?

 

So Wales’ example is spot on and I’ve seen the same syndrome when searching for many, many things. I’m wondering how they’re going to get the community going. It’s different from Wikipedia- when I edit a page on Wikipedia I am supposed to feel like I am adding to the world’s body of knowledge. When I list a few hotels I’ve stayed in I feel like just another consumer talking to random consumers and thinking about it now I am totally unmotivated to do that.

For the moment, by the way, I’d recommend using del.icio.us for searching for anything that’s likely to be spammed. It does a pretty good job even though it’s so simplistic.

 

It looks ironically like Google.

 

Interesting. Though this will intensify spamming of Wikipedia as webmaster’s seek to become a “trusted resource”.

 

Wales does correctly identify the problems with spam on Google. I also find a lot of problems searching for something and getting pages of review and price comparison sites, which actually don’t have any review or prices on them. In an interview Google said they knew that happened and it was completely fine - a legitimate candidate for being in the index.

Jason is spot on - I’ve recently started searching del.icio.us for things and following the ones which have been bookmarked the most with fairly good results.

Maybe what we need is something with Digg type tools to promote or bury a site in the SERPs? You’d soon “digg down” a spam site, and give a thumbs up to a good one.

 

I wonder about the screenshot; it appears to have come from this site:

http://www.searchme.com/beta/

Despite the Wikia badge at the bottom, there’s not an unambiguous statement of affiliation between ’searchme.com’ and Wikia.

Also, the ‘Wikiasari’ name dates back at least 2 years; see it mentioned in this November 2004 page revision at Wikia, back when Wikia at first appeared to primarily be a wiki-style DMOZ rather than hosting:

http://search.wikia.com/index......;oldid=658

 
 

I doubt it could really work well, as there are too many keywords for search. Digg’s spam tool works, but how could a tool works for millions of pages?

Tech Tutorials: http://www.hotcoding.com

 

Huh?

How about they fix the search on Wikipedia first. You need Google to find anything in Wikipedia. Same with Amazon. How hard is is to implement a “Did You Mean?” feature! Amazon should pull those eggheads at A9 over to the main site and get the search working.

 

A Digg’s spam toll works as long as people don’t have any problem with it. Controversial comments and so on is always dugg down because they don’t agree with what is states even though it can be true.

 

And here has been me thinking for months that Wikipedia desperately needs Google as it’s search engine. Current Wikipedia search is useless, can’t do that one vital thing Google has of recognising that the inputted search term may not be 100% correct i.e. different spellings for the same entity or a basic spelling mistake.

 

This makes total sense. Most of my searches nowadays end up at Wikipedia. Why don’t we just bypass Google. Also wikipedia’s references on a particular topic are peer-reviewed and on-topic so not only do I get an explanation of what I’m looking for but I also get the best references and not have top bother “screening” Google search results.

This ofcourse is based on the fact that you are looking for a particular concept/person/topic.

Cheers!

 

Wikia does not own Wikipedia. Wikipedia is not involved in setting up this search engine; Wikia is an independent venture. Wikia is free to use Wikipedia content, just like any other company, as long as they follow the license terms.

 

The screenshot has nothing to do with anything related to the Wikia Search project, which in turn has nothing to do with Wikipedia.

 

Why are we even allowed to write comments on posts.. useless

 

Therer is a certain amount or a great amount of intertia with a market leader such as google.
It is tue that the Google product itself was spread successfully by “word of mouth” as being a better product as a search engine than the market leader at the time which was Yahoo.
Yahoo still commands a good market share out of intertia and familiarity by formet users who know the system are familiar with it and are lax to change. Change is never an easy thing.
Google has such a similar advantage with its users.
As well Google has not been standing still and knows the lesson that a superior search engine service whose company has little marketing clout and presence can displace it.
Still Google is such a dominant market force it is hard to imagine it being superceeded .
http://www.vintagecomputermanuals.com

 

Hey…I hope these guys have their “marbles” together and are not getting carried away. If Google cant find it then maybe…just maybe the results are not there.

This to me is branching out into an area where there is dominant leader (not even Yahoo can keep up) without any revolutionary technology that will somehow show that they will be able to do better than any of the Engines out there.

Cant understand the thinking…..

http://www.jollyjo.org

 

It would be interesting to know how will they combat the spam once the search feature it out. Wikipedia is still relatively clean from spam these days, but the search feature may draw more spammer in for the profit.

 

This is a Google killer, ads by Google!

 

I think it is pronounced, “Wikia, sorry.” As in, I know you wanted either Google or Wikipedia, but your got Wikia, sorry.

 

If they have any intelligence, they’ll ditch the wikiawhatever name and choose a simpler one. Dangit. Even wikisearch is better

 

Are you planning to update the post, Mike? Jimbo says the screenshot has nothing to do with Wikiasari.

 

Wikiasari = DOA as a SE. However, it could be an excellent replacement for their current onsite search system.

 

Wikipedia taking on google? What a joke???? Is anyone else seeing this disgusting spam photo here http://en.wikipedia.org/wiki/Main_Page ???

 

Rumor has it that the Wikiasari engine comes from a start-up called searchme.com, formerly Kavam, funded by Sequoia Capital, the same guys who backed Google and Yahoo - there is big time money behind it.

 

site sounds like it would beg to be gamed

 

> Try searching for the term ‘Tampa hotels’, for example, and you will > not get any useful results

Yep, that’s when you can turn to Yahoo!

 

> Try searching for the term ‘Tampa hotels’, for example, and you will > not get any useful results

Yep, that’s when you can turn to Yahoo!

Really, try it.

 

Well, ‘Tampa hotels’ brings up a bunch of sponsored links, such as Hyatt and Marriott, and a bunch of normal links to other hotels and travel sites. I wonder what is not useful about that.

IMHO the power of a search engine is unleashed when the searcher knows how to formulate a good query. Maybe ‘Tampa hotel guide’ brings up better results. There is even an illustrated book, “Google: The missing manual”, for those of us with less than one working neuron.

 

Surely something like del.icio.us would be better for something like this?

 

Have you checked how many people have registered wikiasari domains in different countries?
Strange that these guys haven’t secured their domains before publishing the news!

 

And even stranger considering that they didn’t secure wikisearch before…
Same mistake twice!

 

Its need of time ..
I think its in interest of common surfer to have more variety and competition in search engine niche to apply proper checks and balances over big players. :)

 

I hope that screenshot is a fake. Otherwise, it’s just another search engine living in Google’s shadow. What’s the point? Most one-word searches on Google turn up Wikipedia entries anyway.

 

I’d be curious to see what measures they’ll take to keep spam and marketing departments out. What’s to stop Toyota marketing from going in there and editing the entry to remove negative press, or put their preferred advertisements as top links?

It’s a problem that already exists with Wikipedia and I don’t think they do enough to stop, particularly when it’s done subtly. (I remember I was reading the article on “Dandruff” a while back and stuck right in the middle was the sentence “The gold standard for Dandruff prevention is yadda yadda, found in Selson Blue”

If this thing ever takes off, the problem would only be compounded. With enough money at stake, there will be some serious attempts at gaming the system - just look at the problems that Digg has.

 

The current switching cost of a search engne customer is zero.

 

I hope this fails as soon as possible. I will never use this, just like I never use Wikipedia, because I *know for a fact* that wikipedia is run by a bunch of biased elitist censors, and so will this new thing. Try to edit a popular article on Wikipedia and see how long your contributions last, or for that matter how long *you* last.

The results on Wikipedia and Wikiasari are only as good as the lowest common denominator of who contributed them. Not a very confidence-inspiring thought… At least machine-based search has some algorithms that can be understood and predicted. Yes, they’re open to manipulation by SEO, but so what — algorithms can be improved over time, people — not.

 

Eric I have a more basic question — why *should* anyone want to keep Toyota’s marketing department out of the SERPs? If they provide useful information about their cars, why does it not deserve to be listed?

This idea that oh no, god forbid, we’re seeing a ’sponsored page’, our brains will turn to mush and we’ll just whip out the wallet and spend all our hard earned cash on a scam… C’mon, most people have more brains than that. For the rest of you, you should have stayed in kindergarten and continue to play with construction blocks.

 

This will never work, haven’t anyone heard of DMOZ.org??? WikiSearch is NOT wikipedia. Google allows you to find things that you will never find on this WIKISearch. I don’t want to find easy things, I want to find the strange things.

 

It is intriguing to find ‘Sponsored Lists’ as part of the screenshot..

Has Jimmy Wales finally started earning???

 

Actually you do get useful results searching for ‘Tampa hotels’. About the only thing google doesn’t do well is searching for c/c++ code, since it cuts any special characters out of your “search string”. That’s the only thing any search engine could do that hasn’t been done.

In fact, Wikipedia’s search page (that you get when your first search missed) doesn’t work half of the time. However if I copy my search term to the Wikipedia search box in firefox it works…

Maybe you could fix that first Jimbo

 

Wouldn’t something like this be bad, people could still find a way to game the wiki system. I mean if Amazon is an investor, surely they’ll expect some of there pages to be ranked as “high quality”.

 

…and formulas… why can google tell you the answer to 2 + 3, but not give you search results for something like x^2 + y ^2, or O(n^2 log n) ?

Certainly wikistupidname isn’t going to be able to do this either.

 

Jatalla.com is already doing 100% people-powered search: http://www.jatalla.com.

As Jeff said above: “The real answer to search and most likely the next google will be people powered search…. It’s true that people are the best source of knowledge, machines are good at quantity, not quality.”

I agree.

 

Hmm will this make my Wikipedia stocks on trendio go up? http://www.trendio.com/word.ph.....anguage=en

 

#13 & #63 are on the money.

DMOZ has been at this for many years and it sucks.

I use wikipedia on rare occasions…A year or two ago I tried adding an external link to my site within a certain article…If users were reading this article, they would surely be interested in my site - it was a very relevant link IMHO. My link was deleted within 24 hours by what can only be called the wikinazis.

There are big, huge, massive problems when you allow people to control content - just look at digg & dmoz…the systems are gamed. With DMOZ - if a competitor is the “editor” of your category - forget it…you will never get listed. And with Digg - there are only about 100 people who get stories promoted to the homepage.

It just doesn’t work…or at least…no one has found a way to make it work - yet.

 

that reminds me of another search engine i’ve seen. it starts with a g….ends with an e, i want to try it though

 

read this on http://www.wikiasari.com

Reporters and bloggers note: Amazon has nothing to do with this project. They are a valued investor in Wikia, but people are really speculating beyond the facts. This has nothing to do with A9, Amazon, etc.

Help me out, spread the word. I am looking for people to continue the development of wikiasari. Specifically community members who would like to help build people-powered search results and developers to help us build an open-source alternative for web search. Discuss here. Join the mailing list. —-Jimbo Wales 23:24, 23 December 2006 (UTC)

Update: The TechCrunch story is also wrong. This project has nothing to do with the screenshot they are running, and this search project has nothing to do with Wikipedia.

 

It states “Search Wikipedia and its external Links”.

I hope that will make some editors at wikipedia loosen up a little and actually listen to discusions about what link to and what not and not just settle for a link to a DMoz or Yahoo! Directory Category.

Jimbo and his Wikiasari Project do not benefit from such attitude. Quite the oposite. It would hurt the Wikiasari search Result Quality.

 

“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.”

BZZZZT! Wrong. This search engine _will_ fail. There is no such thing as “good” or “bad”. There is only appropriateness for a particular situation. Since there are many many situations, each page has many different “goodness” or “badness” scores. You can’t just “build a community of trust”. It’s orders of magnitude more complex than that.

 

Look at the name: Google. What does it mean?
The web volume is exponential growth. It’s changing by microsecond.
We can’t go back, period.

 

Pretty unfair game, Google helped a lot (keeps helping) of Wikipedia in getting to the point where it is today, mainstream popular site… In my view more than 50% (I’d bet on 70%) of all searches performed on Google are organically directed to Wikipedia’s web properties…
My 2 cents

 

its ironical to compete with G by using their results somehow!

 

This will put TREMENDOUS pressure on wikipedia. If this search engine takes off, it’s going to require the amount of (FREE, UNPAID) work by wikipedia editors to expand 10 fold.

This has the potential to ruin Wikipedia more then it already has been by spammers, etc. And it means that getting legitimate links into wikipedia will be that much harder.

I am not a very active wikipedian, I edit perhaps 10 articles a month and have contributed about 10 new articles, but I can see that such a project may be good for Jimbo, but it’s not good for wikipedia.

I want to edit wikipedia so I can help spread knowledge, not to A. help enrich others or B. get into endless wars over what links should be allowed and what shouldn’t.

Perhaps Wikipedia will have to outlaw outbound links because of this? Otherwise every day will be a battle.

 

Yes we all know that google can and does produce some crap in their searches, but I am not sure how Wikiasari is not going to fall victim? As far as the community aspect goes, just can’t see it being free from corruption. Just ask DMOZ or someone who has tried to get a site listed. I am not saying that all DMOZ editors can be bought, but now you have the HUMAN factor. The HUMAN factor will be no different than the Computer aspect of delivering search results.

 

I’m staying with Google, and leaving Wikiasari to people who need the benefit of human-edited stuff.

The latter is a huge market.

 

Wow, the screenshot provided looks almost exactly the same as google. The page layout is even pretty much the same. I wonder if they will run into copyright issues.

 

is wikiasari investment worthy? I have a poll for such a proposition at http://japannai.blogspot.com/

Tyler

 

THE SCREENSHOT IS REAL AND YOU CAN BECOME A BETA TESTER AT http://www.searchme.com/beta/

 

This is Search 2.0. I am launching a Digg like wiki, http://www.tagbud.com.

 

TechCrunch Japanese アーカイブ » Wikipedia、独自の検索エンジンを準備中―独占スクリーンショット

Bee Zebb: GalaxyGarden.org

2006年 12月 27日 at 3:03 am

Most Excellent News for the Development of an An Alternate WebBrain to the Growing GoogleBrain….

Now dive in, my WebMonkeys & generate intel & eyeballs, for thee Future depends on your attentive details….

BzB in the GalaxyGarden.org (a GoogleBrainer)

Mauna Kea MotherShip!

 

We’ve already begun a similar project at Bessed ( http://www.bessed.com ), allowing visitors to freely add results for search phrases via comments on each search, BUT…

…we are using PAID human editors to make final determinations on rankings and sites included, with visitor suggestions added on, and other site visitors getting the chance to give yea or nay opinions. (The site’s built on WordPress to allow comments on each topic/keyword phrase.)

Our problem will be scalability; Wikiasari’s problem will be getting volunteers to help create a search engine with the knowledge that half the profits are going to Jimmy Wales. Wikipedia works because it’s useful AND no one feels like they’re working for free for the profit of others. DMOZ didn’t work because too many volunteer editors have their own agendas.

Human-powered search is a great idea–but one of us has to make it a win-win for it to succeed. Pay your editors and the site builds slowly; ask them to work for free while you profit you’ll get abuse of the system.

I don’t know that Bessed has the answer, but I prefer the way we’re going about it to any of the other human-powered attempts that have been created to date.

 

>

The first unpaid Google search result on “Tampa Hotels” is TripAdvisor’s Tampa page. TripAdvisor is a good site that consists of user-generated reviews and a user-managed wiki.

 

@ Robert,
The screenshot is NOT real. Its even mentioned on the search.wikia.com page !!

However, YES, you can sign up for the Beta.

 

The screen grab is very real but it is a screen grab of WikiSearch (a different Wikia project) NOT Wikiasari.

Wikiasari (probably not what it will be called) is Jimmy’s vision for a community powered search engine. It has not yet been developed and is still in the concept stage - it is also not a commercial product - and will definitely be open source.

WikiSearch (what you saw in the screen grab), on the other hand, was developed by and is a trademark of Searchme, Inc. using a totally new search technology proprietary to Searchme (it is a search engine for Wikipedia and External Links). WikiSearch will launch in mid-January. SearchMe and Wikia have a contractual agreement to donate the majority of the revenue from Adsense ads on WikiSearch to the Wikimedia Foundation. WikiSearch is a commercial product created to raise funds for the (non-profit) Wikimedia Foundation with the approval and cooperation of Wikia. WikiSearch is community driven in the sense that it indexes only pages external to Wikipedia which have been selected and approved by the community as External links — but it is not what Jimmy is envisioning for the future.

Randy Adams
CEO
Searchme, Inc.
randy@searchme.com

 

But what about commercial listings. Will this new Wikia search engine contain any commercial sites since most of wikipedia doesn’t.

 

Compare the screenshot with this text from google: Results 1 - 10 of about 1.820.000 for toyota camry. (0,15 seconds) .
DO you wanna compare?

 

Wasted white space. Seems everyone lays out the page to please the first-timer’s eye at the expense of making the ads and results more usable. Look at all the wasted white space in the top 25% of the page and then there’s room for only 2 search results after all the ads. 50% of the second and third lines each is wasted in the right side sponsor ads because the text does not run all the way across the area allotted. Everything in green is a waste because no one, NO ONE, wants to see a www thing when a click anywhere will take you to a site. Nowhere can an advertiser put a pic or logo on the search screen at any price. Offer them a choice between Google and Google and they’ll pick Google every time.