Wikia Search lives up to expectations.

Posted on 7th January 2008 at 9:52 am by Manley

Wikia Search, the new search offering from Wikipedia founder Jimmy Wales' for-profit company, Wikia inc., is now in alpha.

Will this project prove to be the competitor to Google which Wales predicts?

New search engines appear all the time, in a number of guises.

• Many are meta search engines, aggregating the results from other engines and providing results which are based on a more specific user model, be it audio, video or any other niche. Famous meta engines include Copernic, Kartoo or Dogpile;
• Some are trying to build a new model for interpreting user behaviour data, such as Trexy.com, which allows users to 'blaze trails' and records similar end points for searchers of similar terms, or ChaCha, which has human search guides to aid with your search and is able to record the results which are most useful to the end user.
• Some, such as Wikia and, to an extent, Yahoo! Answers, are using direct human editing to provide a truly social search platform.
• Directories can be considered to be a search engine of sorts (including Yahoo!, ODP and , to a certain degree, Wikipedia itself).
• The final major search providers are the tag engines, such as Technorati and Bloglines, although there are plenty of personalised verticals which have their own dominant brands, particularly in legal and medical fields.

Wikia appears to be aiming to combine the human input of social bookmarking sites, such as del.icio.us, with the directory aspects (with the 'Mini Article' feature) of Wikipedia.

I covered the proposed launch of this engine, although back then it was to be named 'Wikiasari', for a presentation back in January 2007. At the time Jimmy Wales went on record as saying:

"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 judgements, 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."

My response at the time was quite damning of both the project and the market:

"Wikipedia has a very strong following, but that there is a strong tendency towards didactic censorship and required notability which is increasingly commented on and which will detract from the final results, and I think that users are rapidly losing confidence in Wikipedia which, coupled with its already appalling search reputation (most people search Wikipedia through Google), should doom this project to failure.

All in all, it seems to be an over hyped bad idea. I have no doubts that it will take off."

The reality is that I would not be reviewing a new-launch search engine at all, were it not for the fact that Wales is behind it. The success of Wikipedia has meant that Wikia is big news and it will get a good number of users. My reasons for looking into the site and my previous reservations aside, I went to have a look at how it performs.

Now, this is an engine in alpha, so if you go there today and look at the results for any given search then you will be disappointed. I am not setting out to compare this engine with Google; the gulf is too wide and I may as well try to compare the Armstrong Siddeley Lancaster to a Yamaha YZF600R. There are certain elements which are the same, but essentially they are different beasts altogether.

Despite this, Wikia does not perform as well as I would have liked. As yet it is chock full of spam and the peer review model does not seem to be up to filtering it out as yet.

A prime example is that a search for [viagra] is currently returning Matt Cutts at #1. Presumably this will soon settle down and, with such a high profile page, will be amongst the first clean-up jobs, but if Wikipedia's record for accuracy is anything to go by (and they have the birth date of Jimmy Wales wrong on his biography) then Google probably do not have too much to worry about.

Do not get me wrong, I am not a fan of human powered search I believe that algorithmically harvested user behaviour data is far more accurate than that supplied explicitly by the users themselves, but I am not writing this experiment off, as some of the industry seem to be. The open algorithm will initially breed huge amounts of spam, but eventually it could become an excellent resource.

For now though, that 'eventually' is years away and is only a vague possibility, just like all the other start-up engines. In order to break into the mainstream search market an algorithm needs to be something really, really special and I am just not seeing this at Wikia yet. I hope that Wales can make a go of it - competition is always welcome - but for now Wikia is just another start-up, albeit one with a lot of hype surrounding it.

As with Squidoo in 2005, or Knol, when it arrives, I will experiment, prod the security, add a mini-article or two and then just keep an eye on it and see how it matures. Wikia is not going to be changing the search paradigm for a few days yet. The launch of Wikia has filled the quiet winter period for search, but ultimately it is a bit of a non-event.

[UPDATE: As predicted, Matt Cutts has disappeared from the Viagra page, a mere 45 minutes later. Hopefully this is a portent for a successfully monitored engine. I shall keep watching.]

File under: wikia

Permalink

Comments

No-one has commented so far, or all comments are awaiting moderation.

Post Your Comment

Subscribe

RSS 2.0 Feed