Wikia Search Upgrades Get Closer

Barence writes "Wikia Search has been revamped with a wealth of editing tools, to build on its early promise of being a more open search engine. The firm behind Wikipedia has introduced a wealth of editorial tools allowing users to directly edit and annotate results with text, images and links, as well as spotlight specific searches or even delete a site from the search results, a change which affects everyone who conducts that search in the future." Update: 06/04 17:32 GMT by T : Jimmy Wales wrote to point out that despite his role in founding Wikipedia, Wikia is a completely separate company.

Comments

"Open"

[Anonymous Coward]

delete a site from the search results, a change which affects everyone who conducts that search in the future

So it's possible to define censorship is "open" now? Awesome.

And, Yes Yes. I know that one can go to some effort to determine whether a result has been censored and Undo It Yourself in the manner of Wiki. But how is it that you need to have expert knowledge in order to get uncensored information? Is that not the definition of elitism?

Re:"Open"

[iago-vL]

You're assuming that it'll be censored and say that way. I doubt that'll be the case -- if a user "censors" something, just like when a user vandalizes Wikipedia, it'll likely be reversed fairly quickly. It would surprise me if this became an issue.

Re:"Open"

[nahdude812]

Except that we know many admins on Wikipedia have repeatedly abused their position to advance a point of view on a topic - sometimes even for a topic they have a vested personal interest in - and suppress an opposing view.

Who's to say that when an admin of Wikia Search also owns one or more sites in a given category, they won't use their power to repeatedly squash their competitor sites? And that when it comes up for review, their you-scratch-my-back-and-I'll-scratch-yours admin friends won't be there to defend them?

Jimbo needs to do a lot of work in regard to increasing visibility of administrative process, and addressing administrative abuse (his typical response has been to blindly defend his admins).

If he doesn't, abuse will become a lot more rampant, a lot more widespread, and a lot more subtle since now we're talking about something which can directly affect profits of a company and those of its competitors.

Closer to What Exactly?

[eldavojohn]

Wikia Search Upgrades Get Closer Closer to?

Wikia Search moves closer to open ideal Oh, an "open ideal?" Is that a problem? I thought that was all Wikia's Search had going for it right now.

Don't get me wrong, that's about all Wikipedia had going for it at one time. That and a meager couple thousand pages.

But I think Wales' search project should aim a little higher at this point in time ... the improved tools are great and I don't doubt this is the openest search engine out there ... but shouldn't we all be a little more concerned about how well it works?

Also, maybe the reward system should be a little more tangible, like Wikipedia's? I mean, people get a kick out of seeing something they write benefiting everyone in the world. I personally don't get a kick out of knowing that if someone searches for 'Lola by The Kinks' it now comes up with more accurate results. I question the rewards although, to be fair, I also questioned the rewards of Wikipedia.

It shall be interesting to see whether or not this takes off. I encourage Wales to keep trying because even if this doesn't work, it certainly hasn't cost me anything. I wish him and his staff the best of luck!

Re:Closer to What Exactly?

[saibot834]

Wikia Search needs time. Like Jimbo says, when he started Wikipedia he called it an encyclopedia, even though at the time there were no articles. Wikia Search is still Alpha and people should stop thinking its a finished or even practically usable search engine. (Here is an interesting conversation with Jimbo (read the comments))

BTW: There seems to be some confusion about Wikia and Wikipedia. Short info:
* Wikipedia is a free encyclopedia that uses MediaWiki and is run by the Wikimedia Foundation
* MediaWiki is a popular Wiki-Engine
* Wikimedia Foundation is a non-profit organization that runs Wikipedia and several sister projects
* Wikia is a for-profit organization that has _nothing_ to do with Wikipedia _or_ the Wikimedia Foundation! One of their projects Wikia Search; but they also host many Wikis. The only thing Wikia has in common with Wikipedia is that it uses MediaWiki and was started by Jimbo Wales.

Re:Closer to What Exactly?

[sm62704]

Wikia Search Upgrades Get Closer

Closer to?

Perhaps "closer to actually being useful?" I hadn't tried it before, but clicked the link to it from TFA.

I wanted to find the aging but still humorous "police warning" about the new date rape drug "beer". So I luugged in "date rape" beer. None of the first page results returned a copy (there must be hundreds) of the document I was looking for.

Google had the one I was looking for at the top of its list. Wikai search still has its work cut out for it. A search engine that can't beat Google or add something to the party really has little reason to exist.

Re:Closer to What Exactly?

[patro]

Also, maybe the reward system should be a little more tangible, like Wikipedia's? I mean, people get a kick out of seeing something they write benefiting everyone in the world. I personally don't get a kick out of knowing that if someone searches for 'Lola by The Kinks' it now comes up with more accurate results. I question the rewards although, to be fair, I also questioned the rewards of Wikipedia.

Well, maybe the reward could be search results which YOU find relevant.

I think Wikia Search could be even more revolutionary than it is now. How about dumping the idea of a common search result page which is the same for everyone and displaying results which are relevant to the individual user?

Here's how it would work: people can rate the search results and let's assume they do. If they do then the system stores how they voted in the past. When displaying search results the system would check who else voted similarly to me in the past and would calculate the result page using that information giving more weight for the vote of those people who share my tase.

The result of this would be that I would see a result page according to my own taste. I would never see Lindsay Lohan, because I and the people voting similarly to me would vote it down, so this ad-hoc loosely-coupled group helps its members to avoid news about Lindsay Lohan.

Christians would vote down porn, so they would see only Christian-approved stuff. No need for any kind of censorship, because their group takes care of that. On the other hand, porn lovers would vote up juicy sex movies, and vote down religious propaganda, so they also would see only what they want to see.

Spammers would have no chance, because their material would be voted down instanly by me and my peers, so I wouldn't see them, but the spammers themselves would, because they voted it up, but noone else would see it, but them.

The idea of the denominator search results should be forgotten, and we should get individually tailored result pages instead based on our voting history.

Needs Work

[Anonymous Coward]

I did a search for "magazine article search" (no quotes) on Wikia Search and the number one result was a Wikipedia entry about Jack Kemp. It did better if I included quotes. Their ranking algorithm seems to need work.

Problems?

[Archangel Michael]

as well as spotlight specific searches or even delete a site from the search results, a change which affects everyone who conducts that search in the future." Anyone else see anything wrong with this?

I want my results to be as agnostic as humanly possible, which means keep human hands out of my results.

The problem with this type of search is that those with agendas will control the results. While I may like the agenda or support it, I don't want to limit what I see to just my POV. There are too many people who can't think beyond their own little world.

I don't want Scientologists skewing results. I don't want Bush Bashers skewing the results anymore than Obama Haters. I don't want KKK results being skewed up or down because of an agenda (or two).

I want to see Huffington Posts along with Michael Savage. The moment someone starts monkeying the results, the agnostic nature of searches goes away, and the search will become useless to me.

Re:Problems?

[FooAtWFU]

as well as spotlight specific searches or even delete a site from the search results, a change which affects everyone who conducts that search in the future." Anyone else see anything wrong with this? I want my results to be as agnostic as humanly possible, which means keep human hands out of my results.

Good. Have fun searching spam.

Seriously, even Google's PageRank stuff relies on "human hands" placing links to documents on web pages.

Re:Problems?

[chunk08]

So true. While the philosophy of Wikipedia and Wikia would perhaps hope that this would bring more "relevant" search results, it will be controlled by agendas. Wikipedia is, but so many people with different agendas edit it that everyone pulling in different directions "tends" to move it toward the middle. Since results can simply be deleted in Wikia Search, and one must go to some effort to find them again, the first person with an agenda who edits a search (the first person who edits the search), will most likely skew that search to their bias forever.

Re:Problems?

[MozeeToby]

On the other hand it could make a handy mobile bookmarks system. Just do a search for "aeqoucqoueiriue123" then add all your bookmarks to the results. Next time you want to see your bookmarks just do the same search again.

Re:Problems?

[XanC]

Ah, Dvorak gibberish. I'd recognize it anywhere. Lovely layout, isn't it?

Re:Problems?

[Archangel Michael]

I'm pretty sure you don't know how Page Rank works.

From Wikipedia .....

PageRank is a link analysis algorithm that assigns a numerical weighting to each element of a hyperlinked set of documents The results are based on an algorithm, not human manipulation. While PageRank isn't perfect, it is still about as good as anything else out there.

Re:Problems?

[Kingrames]

The way I see it, it's no more censorship than letting everyone who searches say "that has nothing to do with what I was searching for" and actually seeing the computer respond.

I'm fairly certain that if the people you mention DO engage in dubious practices like that that their accounts will be banned, their efforts undone, and their changes revoked.

Then, in accordance with the Streisand effect, everything they hoped to accomplish will fail and backfire.

Re:Problems?

[Anonymous Coward]

Should be just a matter of having the right policy. Only allow spam to be removed and nothing else. If somebody is found removing what he shouldn't or adding spam, block him for a couple days.

Re:Problems?

[Anonymous Coward]

Seriously, even Google's PageRank stuff relies on "human hands" placing links to documents on web pages. Surely you mean pigeon beaks?

Re:Problems?

[Anonymous Coward]

The results are based on an algorithm, not human manipulation. ...and all this time i thought it was done by pigeons.

http://www.google.com/technology/pigeonrank.html

Re:Problems?

[ThomasLB]

The major players in the computer business seem poised for a wave of mergers, and when that happens we won't have "agnostic" search results anymore. Instead, we'll have advertiser-driven search results skewed to improve the corporate bottom line. In the near future, an open-source community-driven search engine will be a very valuable resource to have.

Re:Problems?

[pablomme]

The results are based on an algorithm, not human manipulation. The GP's point is that the data the algorithm uses is supplied by humans (those who write the webpages that link to other webpages).

Re:Problems?

[owlnation]

I'm fairly certain that if the people you mention DO engage in dubious practices like that that their accounts will be banned, their efforts undone, and their changes revoked.

Well if Wikipedia is the model then, yes that will certainly happen. After all, Wikipedia has blocked entire countries' IP addresses on the whim of Jimmy Wales.

The problem is... define "dubious practices".

Wikipedia blocks for "vandalism". But has never satisfactorily defined what that is -- often in practice it's disagreeing with the cabal that runs a page who is either an admin or has their view supported by an admin.

I see no reason why this will be any different here. Given Jimmy Wales's well documented dubious financial dealings and ethics, no doubt sites like answers.com, wikipedia itself, or other sites where he has a vested interest, will get priority regardless.

Based on the Wikipedia example, I can see no way in which this search engine could ever be trusted.

Re:Problems?

[Archangel Michael]

All Data is touched by Humans, so that is a specious argument.

Data that is massaged by humans to get a certain result is tainted. Which is one of the flaws of PageRank system. I didn't say PageRank was perfect, just about as good as it gets.

The point is, whenever someone learns how to game the system, it will end up like when PageRank gets exploited by irrelevant websites pointing to each other.

I prefer a system where gaming the system is less likely. The approach here is to actually game the system.

Re:Problems?

[pablomme]

All Data is touched by Humans, so that is a specious argument. It's not, it implies that it's not a matter of whether the data is human-provided or not, but of what amount (and type) of human implication is involved.

Clearly WikiaSearch will need anti-spam and anti-censorship measures, but it may well be feasible to make it a good search engine.

The problem IMO is that the quality of the search results is affected by the popularity of the search engine, and the popularity of the search engine is affected the quality of the search results. So for this to work they need a good basic search algorithm to start with. Notice that the success of Wikipedia was possible because there were no other (popular) online encyclopedias available. Not the case of search engines.

Re:Problems?

[pbhj]

I didn't say PageRank was perfect, just about as good as it gets. Then why don't google _just_ use PR to provide list rankings rather than applying adaptations over the top?

I think a moderation system that's less digital would be better.

Re:Problems?

[Ben174]

Upon a quick glance at the recent updates page, it's apparent that Wikia's going to have their fair share of spammers.

Re:Problems?

[GregGardner]

While I agree that allowing anyone to monkey with your search results is ripe for abuse, you are assuming that there are no humans manipulating your search results at the current search engines. While Google loves to tout the fact that their search results are completely algorithm-driven, they actually have teams of people who use manual intervention on a daily basis to "improve" search results, which usually is done to stop abuse of their guidelines.

These people can take actions such as setting a site's page rank to zero or removing entire websites or parts of websites from their indexes. Here is a reference of one such publicized case:

http://news.bbc.co.uk/1/hi/technology/4685750.stm

The fact that you think there is no manual intervention happening probably means that you don't notice it and believe that Google is doing a good job with the manual changes that they are making, but it doesn't mean that it isn't happening. It also doesn't mean that their changes are never influenced by outside forces, as evidenced in another publicized case:

http://searchenginewatch.com/showPage.html?page=2164661

Re:Problems?

[CCFreak2K]

If you take humans out of the equation, this tends to happen.

Re:Problems?

[patro]

"Good. Have fun searching spam."

Spam won't simply disappear if you let humans edit the results.

In case of Wikia Search a spammer can edit the search results deleting the competition and he will be the frist match for a search for a short period of time until his edits are reversed.

If the search engine has lots of users then even this short period of time can be very profitable and brings in thousands of visitors.

I wonder how they will fight it. Delayed propagation of edits into the search results is no good, because then existing spam can also be removed only with a delay.

I envision search wars where different groups fight bitterly for the control of certain search terms, just like the edit wars in Wikipedia.

It's simply unusable

[Anonymous Coward]

I tried a simple search for Google, and aside from looking like shit, the page causes Firefox to really stutter with all the junk that's on it. This is an embarrassment.

can also use google

[troll -1]

Wikia Search also includes buttons allowing users to try their search on the other major search engines with a single click ...

You can do site specific searches right from google by using the 'site' option. So to search wiki for 'lenin' you would enter the following on google:

lenin site:wikipedia.org

Re:can also use google

[Melbourne Pete]

Wikia Search also includes buttons allowing users to try their search on the other major search engines with a single click ... You can do site specific searches right from google by using the 'site' option. So to search wiki for 'lenin' you would enter the following on google: lenin site:wikipedia.org That's two different things altogether. Wiki Search supposedly lets you use other search engines from the same page. The Google example you gave lets you confine the same Google search results to hits from one site.

Re:can also use google

[Anonymous Coward]

O/T Most of the time its quicker to just search for 'sitename "searchterm"'. 'sitename' seems to work as well as 'site:sitename.tld' in most circumstances.

Re:can also use google

[risk one]

That's not what that sentence says. What wikia has are buttons that try your search on other search engines. It's like Google having a button "try this search on Altavista". Not a great functionality on established search engines, but on a system like this, that needs to find it's place, it's a pretty good idea.

The main reason they're doing this, I imagine, is that they want to be people's home page, so it's a good idea to offer the same functionality as their previous homepage (eg. Google).

Re:can also use google

[Anonymous Coward]

lenin site:wikipedia.org And you can search for other beatles too:

mccartney site:wikipedia.org

How about making the current search not suck?

[Anonymous Coward]

Wikipedia's search function for wikipedia as a whole blows. Compared to google, its tolerance for errors is laughable. A lot of times, I'll be searching for something I've only heard once or twice, and have no idea as to the spelling for it, or the complete name. Most of the time, I'll have to go to Google and type my search in to find the actual Wikipedia page for the topic. On the first try, Google finds it. Amazing!

Yeah, they should really work on it.

Wikia does not own Wikipedia

[Anonymous Coward]

Wikipedia is not owned. If any organisation can lay claim to such a thing, it would be the Wikimedia Foundation.
Wikia is a for profit organisation. It aims to do well.
Thanks, GerardM

Wikia is not the "firm behind Wikipedia"

[Anonymous Coward]

Wikia is not the "firm behind Wikipedia", Wikipedia is operated by the Wikimedia Foundation. I know words with "wiki" in the front all sound the same, but that doesnt mean that they're all the same company.

Will use for one purpose

[JD-1027]

Whether I find it useful as a search engine or not, I will find at least some gratification in that I can do nothing but delete all search results from experts-exchange.com

Yes, I'm bitter, I can't count the number of wasted clicks on their search results from google. Now that google cache doesn't work with experts-exchange, it is nothing but a time waster.

Re:Will use for one purpose

[cerberusss]

Huh? If you scroll way down, then the answer is always there in my experience. It's not so nice that it's "hidden" that way, but you only have to scroll down.

Re:Will use for one purpose

[geminidomino]

Nope, now the cache shows the same thing the click-through does.

"All comments and solutions are available to Premium Service Members only.

Start your 7 day free trial and see for yourself why Experts Exchange is the easiest and most proven technology resource in the world. Get Started

Already a member? Login to view this solution."

Gods, my kingdom for a google blacklist

Re:Will use for one purpose

[cerberusss]

Eh? Search for Chmod all files within a directory Linux, then click the first link and scroll down completely. Voila, solution.

Re:Will use for one purpose

[geminidomino]

I dunno if it's your IP or what, but I still see this for the "accepted solution" (and all others)

http://img261.imageshack.us/my.php?image=screenshotof8.png

Re:Will use for one purpose

[cerberusss]

Looking at the slider, I see that you haven't scrolled all the way down.

is wikipedia really trustworthy anyway?

[andre3001]

Haven't they had enough problems with this before? Seriously, there's one poor guy out there who's already spent most of the last year trying to keep vandals out of wikipedia pages for Clinton, Obama and McCain. (http://www.chron.com/disp/story.mpl/headline/nation/5808949.html) And weren't there more issues last year with everyone from major corporations to the Vatican rewriting their pages? Are they just going to employ an army of editors to make sure we're not getting false info? What's the use of information that no one can trust?

Re:is wikipedia really trustworthy anyway?

[risk one]

There's no such thing a perfectly trustworthy information. Ever, anywhere.

The following goes for all information: If you care about being able to trust it, you verify with another source. The more you care, the more you verify.

Wikipedia has a hard limit to its trustability, and within that limit they are doing very, very well.

So in answer to your question: he use of information that no-one can trust is as a starting point. It's the use of any single source of information.

The firm behind Wikipedia?

[Anonymous Coward]

No. Wikia has nothing to do with wikipedia. They are just completely separate entities.

Battle-Merchants

[Mr. Stinky]

When ranking concepts directly related to sellable goods, what should get the higher concern? The sites selling the goods, reviewing, the manufacturer of those goods, etc... I see a strange attractor effect on many people drag-n-dropping their site every minute to the top of the page. I see the potential for serious embattlement over the top placement. I wonder how they will deal with this.

Wikia is mostly fancruft

[Animats]

Despite Wales' hype about Wikia becoming comparable to Google, Wikia is mostly a hosting service for fancruft. The biggest wikis are Star [Trek|Wars|Gate|Craft], the Marvel Comics database, and similar dreck. Wikia's demographic lives in their parents basement.

This is not a basis for a useful search engine.

Re:Wikia is mostly fancruft

[Felipe_Silveira]

Not sure about that. If you consider any good collaborative model (based on mutual benefits) you will prefer fans and geeks rather the regular not-caring folks... If Jimbo succeed when trying to put together the fans of several different kind of subjects, he should be able to build a really neat search engine.... Remember, there are people obsessed about a lot of things beyond Star Trek and etc... And on collaborative model, this weird people are exactly the ones you can rely to freely and happily help you building some great stuff...

Okay, then why the name?

[DancesWithBlowTorch]

Okay, we have all heard ten times now, that "Wikia is not affiliated with Wikipedia". Then, why on Earth does it use this name?

I know, I know, they do wiki-stuff and so on. But I can't help thinking that if anyone else besides Jimmy Wales had come up with that name, he would have a counterfeiting lawsuit at his neck right now.

I'm not sure it's such a good idea either. Forever, it will be "The-firm-that's-not-affiliated-with-wikipedia". I can already see the crowds of wikeeks (wiki-geeks. clever, I know) who'll slowly get horribly annoyed by having to explain this over and over again, once wikia does anything remotely bad (like loosing data, censoring, or some privacy cock-up), throwing a dark shadow on wikipedia.

Re:Okay, then why the name?

[brion]

The term "wiki" and other wiki sites predates Wikipedia by several years; we don't own a trademark on it, nor could we possibly claim one.

There are other wikis with "wiki" in the name, and other companies with "wiki" in the name. There have been since before Wikipedia, and there have been new ones since Wikipedia.

The particular confusion with Wikia primarily arises from having the same spokesperson. :)

Re:Okay, then why the name?

[Frosty Piss]

Excellent. You've made the parent's point. But cry as much as you want, there will always be an implied incestuous relationship between Wikia and Wikipedia.

Be sure to keep mentioning Wikia!

[Anonymous Coward]

The bigger it gets, the more hookers and blow Jimbo Wales can afford!

Re:"Open" ... I CANT wait...

[Anonymous Coward]

A brand new tool from a tool regarded as the King of Tools.

Deleted results are still visible [Re:"Open"]

[Anonymous Coward]

"Deleting" a result is a way of saying "this is result has no relation to the search terms".
Currently, "deleted" results are merely grayed out, they still appear in the results.
In addition, they are sorted to below any results rated with stars.
Thus, "deleted" results are not "censored" in the sense of "removed".
Instead, they are rated low.
Yes, this results in "deleted" search results sliding down, which is the purpose of this feature.

A "deleted" result can be "undeleted" by anyone, by a single click.
No expert knowledge needed there, thus no elitism.

As an aside, are you sure you want to call the ability to undo or edit in a wiki "expert knowledge" (I wouldn't)?