Wikipedia Adds No Follow to Links

netbuzz writes "In an attempt to thwart spammers and search-engine optimization mischief, Wikipedia has begun tagging all external links on its site "nofollow", which renders those links invisible to search engines. Whether this is a good thing, a bad thing, or simply unavoidable has become a matter of much debate." This topic has come up before and the community voted to remove nofollow back in 2005. This new round of nofollow comes as a directive from Wikia President, Jimbo Wales.

Re:Neither good nor bad. It's immaterial.

[Anonymous Coward]

I agree - but I don't think we should be aiming for some eternal game of one-upmanship. There must exist a solution which simply stops the spammers, end of. I am think of something along the lines of a method which includes temporal information in determining things. The big thing about spammers and current methods is all they have to do is figure out how the method works (whether it be page content, meta labels or linkage) and then mimic it to get themselves up. How about if mimicking was impossible - i.e. the kernel of the method is something which operates based on a historical or other time dependent variable. The spam merchants can't go back in time - and if any method they devise take a couple of years to actually get anywhere then they are not going to bother (these folks work on the quick buck, and the amount of time also means the search engine has a very large lead time to change things).

Re:Neither good nor bad. It's immaterial.

[Dan Farina]

Actually this sort of flow model was well documented in IR, AI, and mathematic research for a period long before Google. While credit should be delivered for implementing this scheme in a world of already-entrenched search engines, it falls into the category of age-old computer science. This same scheme is also used to compute the final likelihood of states in Markov models -- a technique at least 30 or 40 years old.

In a nutshell: the eigenvalues of the adjacency matrix.

Re:Neither good nor bad. It's immaterial.

[garcia]

In addition to what you have mentioned above, Wikipedia should not be given the weight it is in Google rankings, period. My Wikipedia user page should not show up as a top five return for a Google search of my name. It shouldn't show up at all simply because it's not as important as the other information out there on me.

The only reason the Wikipedia user entry exists is because Google does rank the pages *very* highly. Bleh.

Not invisible

[truthsearch]

which renders those links invisible to search engines.

Uh, not really. The big search engines choose to not follow those links.

Using nofollow reduces the incentive for spammers, but in this case it will hurt search engines. Google wants to provide the most worthy links at the top of search results. Being linked from wikipedia is supposed to denote reliable sources or very relevant information. Therefore Google is slightly more accurate for having those links to follow in wikipedia. The nofollow will make search engines slightly less useful.

Call this version 1.0

[victim]

This should be considered a step in an evolving policy. The next step should be that old links, ones that have survived many edits and time as well as links added or edited by known and trusted editors should omit the no-follow tag. Then wikipedia can continue to serve as an interpreter of the WWW.

I doubt it'll stop wiki spamming

[jesterzog]

I don't think this will do much to stop Wikipedia link spamming for several reasons:

• Many spam links on Wikipedia aren't commercially motivated spam, but just people who've naively put external links in articles without properly understanding or caring about the editing policy. They're not thinking so much about search engines as about pointing people to their website (or their favourite website) because they think it's more important than it probably is. If it's a relatively obscure article, it might stay there for months or longer before someone goes through and reviews the links.

• Wikipedia is only one of the websites that publishes Wikipedia content. There are lots of other sources that clone it, precisely as they're allowed to under the licence, and re-publish it. They usually add advertising to the content, or use it to lure people to some other form of revenue. These sites are easy to find by picking a phrase from Wikipedia and keying it in to a search engine like Google, and I doubt they'll add the nofollow attribute to their reproductions of the content.

  Wikipedia is probably treated as a more important source of links by search engines, but whatever's published on Wikipedia will be re-published in many other places within the weeks that it takes for the new content to be crawled and to propagate. And links on any Wikipedia articles will propagate too, of course.

• Even if you ignore search engines, having external links from a well written Wikipedia article that gets referenced and read a lot is probably going to generate at least some traffic to a website. Wikipedia articles are often a good place to find good external sources, probably because they get audited and the crappy ones get removed from time to time. This is exactly what provides motivation for spammers to try and get their links added, though.

Good on them for trying something, but I don't think it'll stop spammers very much.

Re:Idea for a New Search Engine with Unique Rankin

[interiot]

Yahoo Mindset [yahoo.com] lets you search for sites that are more commercial, or more informational. Sites with the most nofollow incoming links may fit into the "more commercial" group. (by the way, does anybody know how Mindset actually works?)

Re:Call this version 1.0

[Kelson]

The next step should be that old links, ones that have survived many edits and time as well as links added or edited by known and trusted editors should omit the no-follow tag.

I like this idea. nofollow is more useful for the unmaintained or rarely-maintained site. If you're going to leave the site alone for a month and come back, you probably want to avoid rewarding the comment/wiki spammers who drop by in the meantime. On the other hand, once you verify the site, it's worth helping the site out a bit.

With blog comments, this can (usually) be done through manual moderation: give the links nofollow until the comment is approved. With something the size of Wikipedia, it depends entirely on how popular the target article is. Frequently-visited articles are more likely to have the spam cleared out, and less likely to benefit from nofollow, as it's unlikely that too many search engines are going to drop by in the 15 minutes between the linkspam being posted and the edit being reverted.

An advantage of the criteria you suggested is that it could, in theory, be done automatically. The metadata is already there: how long it's been since the link was added, who added it, how many edits have occurred with the link staying present. IIRC there's also a concept of reviewed/approved versions of an article, where someone has gone through and said, "Yes, this version is good," which could also be used to determine "good" links.

Re:Wikipedia and Internet-Topology

[David Gerard]

"If somebody were really intent on "overgrazing" wikipedia, automated troll-bots would have no difficulty spewing crap all over it faster than the community could work to revert it. I'll be honest, I'm surprised I haven't seen more if it already."

You will be utterly unsurprised to know this happens already ...

In general, any obvious objection to the idea of a wiki encyclopedia already happens and is already dealt with day to day. We have a ridiculous array of spambots and vandalbots already attacking Wikipedia and trying to turn it to their use, never mind our work trying to write an encyclopedia. So we have an EQUALLY ridiculous array of antivandalbots to deal with these things as needed. Our immune system is quite frightening to contemplate at times ...

Re:Wikipedia and Internet-Topology

[David Gerard]

"Personally, I'm astonished that Wikipedia hasn't done this from the beginning."

All the Wikipedias other than English have had this in place already. It's just that the flood of spammers has been so bad on English Wikipedia we've finally had to put it on there too.

Could be a tax issue for Wikipedia

[Animats]

Wales' behavior may be an issue for Wikipedia. If the same person is involved with a profit-making venture and a nonprofit in the same area, the tax status of the nonprofit becomes questionable. When a US nonprofit files their tax return, they have to list any officers or directors involved with profit-making ventures in the same field.

The IRS is concerned because if you have a nonprofit and a for-profit organization under the same management, it's often possible to structure things so that the for-profit corporation shows a phony tax loss.

Re:Wikipedia and Internet-Topology

[frankie]

The current problem with Wikipedia is more of an offshoot from Tragedy of the Commons. In the grand tradition of Slashdot analogy-stretching:

• Wikipedia is the field
• long-time users are the (semi-enlightened, self-regulating) farmers
• HOWEVER, thousands of new farmers have arrived in town, with more every day
• AND it turns out that at least half of them are actually human-shaped insects a la Mimic [imdb.com] trying to devour the field AND the cows

In all seriousness, Wikipedia has simply outgrown its youthful innocence, just as the Internet did about 15 years ago. Peer-reviewed anarchy breaks down after a sufficient quantity of greedy scumbags show up. Semi-protection needs to become the default.

Could they not do it smarter?

[mi]

For example, auto-add the "nofollow" only to the links added in recent edits (for some definition of recent). Once a particular link was part of the page long enough (and survived other people's edits), it can be followed by the search engines...

I, for one, contributed a number of wild-life pictures to Wikipedia, but am also selling them in my own shop [cafepress.com]. I don't think, it is unfair for me to expect links to my shop from the contributed images to be followed...

Wikipedia administration, and no, this is not good

[TheSpoom]

IMHO this is part of what's wrong with Wikipedia. They claim to be open to all and to have a community, deciding many things by consensus.

Except when Jimbo, or another well-known admin overrules everyone else.

They've even sneakily formalized this policy in renaming Votes for deletion to Articles for deletion, suggesting that while a discussion can take place about an article's fate, it can generally be ignored if an admin (typically the one placing it up for deletion) disagrees.

There's some interesting information over at WikiTruth [wikitruth.info] about this (like everything else, taken with a grain of salt; there's some obvious bias there).

Anyway, I personally believe this is a bad thing for the overall health of the internet. Wikipedia is a huge site. Making it irrelevant to search engines will probably affect Google quite a lot, and give a *huge* boost to whoever figures out how to get around the nofollow restriction.

Re:Wikia is not Wikipedia - please correct story!

[EvanTaylor]

Sadly this comes from always being taught to expect what you read to be true; at least in encyclopedia/textbook form.

Not a single history book in any school in America is correct. There are blatant bias lies in all of them, period. Math textbooks have many mistakes as well. Historical fiction may give people the wrong impression or state the fact incorrectly. Just about everything we see, hear, or read is false in one way or another.

Sorry the burden is on you, and you alone, to figure out what is and is not correct. We disprove theories on physics every few hundred years, biology every few decades, history every day it seems. Get over it, learn to think for yourself.*

*Everything stated in this post may also be false, or it may not, please consult someone who only tells the truth, if you can believe them, to see if it is or is not correct. Then again, maybe it is true. I dunno.

Re:uncommonly tragic?

[Anonymous Coward]

If any site has "no-follow" links on it, that means that not only are they not verifying the quality of the links I might click on should I choose to go to the site, they have a general assumption that those links will be bad links. That being the fact of the matter, I already don't want them to show up in my search results at all.

Someone puts a link to a 'good' site on a wikipedia article. Later that domain is sold/expires/the link was just temporary/etc and now (say 3 years later), the link is to a spam site. How do you manage going through every external link in Wikipedia every week to verify that they're still pointing to the resource that was originally intended? Even with rel=nofollow this will still happen, but search engines won't pick up the ++ for having that link pointing to them.