Wikipedia editing gains trust but not necessarily accuracy
Posted by Sandra Cosser on 04 Sep 2009 | Tagged as: Search Engine News
The world
Good news for all of us who live and die by Wikipedia – the ultra-convenient source of dodgy information is finally giving users the means to separate iffy content from that which by common consensus is accepted as correct. What that means is that you’ll be able to tell what the masses think is genuine, but it won’t necessarily tell you the truth. And validity is skewed towards regular Wikipedia contributors as opposed to first time writers who really know their stuff.
According to the website in question (i.e. Wikipedia), “WikiTrust is a MediaWiki extension that implements an author reputation system, and a text trust system, for wikis. WikiTrust adds to a wiki a check text tab that enables any visitor to check the author, origin, and reliability of wiki text. Thus, visitors can easily spot spam, surreptitious changes, and information tampering.” (Emphasis author’s own)
According to Wired.com, this translates into a colour coding system that codes every word according to author reliability and time spent on the site. Content from authors who have yet to establish their credibility will be highlighted with an orange background, as trust/credibility is gained, the colour gradually fades to white.
Basically, the principles involved in gaining WikiTrust fit right in with the principles to which Wikipedia adheres, those of public opinion, collaboration and ego. Virgil Griffith, a Wikipedia software developer and neuroscientist from the California Institute of Technology (who was not involved in the project) calls it the ‘Darwinian nature of Wikipedia’ – the more accurate the content, the longer it will remain unchanged.
As far as I understand it, this means that if someone in the field of biology, who regularly contributes to biology related pages, decides to edit a page on Elizabethan history, she is likely to receive a higher trust rating (colour coding) than someone with a doctorate and 30 years experience in Elizabethan history, who has only recently written her first Wikipedia page. Even if the history buff goes back and fixes the biology queen’s mistakes, they’ll still be painted neon orange and will remain so until enough time has passed (or enough people validate her content) to ensure that the buff wasn’t telling porkies.
I’m not the only one who has issues with the concept, although experts have focused on probable technical hiccups. Ed Chi, a computer scientist for the Palo Alto Research Group, is concerned about Wikipedia’s capacity (in terms of processing power and disk space) to run the programme in real-time. He’s also said that he’s not convinced that Wikipedia fans will find the application useful. This is because it hasn’t been widely tested among users.
But Wiki Lab researchers at the University of California, Santa Cruz say that they are working on methods to make the programme more efficient. They’ve also made every effort to ensure that application is as unobtrusive as possible. Currently, only registered users have access to the ‘trust info’ tab that activates the colour coding system, but when it goes mainstream all users will have to click on the tab before their screens are awash in a sea of trust.
The most important thing to remember is that WikiTrust measures consensus and not truth, something that leader of the WikiTrust project, Luca de Alfaro, is very happy to admit. “If 20 people are all biased in one way, our tool does not know it,” says de Alfaro.
Or, if you prefer Virgil Griffith’s interpretation, “Everyone’s injecting random crap into Wikipedia, and what people agree with more often sticks around. Crap that people don’t like goes away.”
