Tag Archives: California Institute of Technology

Who’s behind Wikipedia: Virgil Griffith’s WikiScanner investigates

Questions continue to be asked of the credibility of information on Wikipedia, but the online encyclopedia is increasingly becoming a tool – and often a first point of call – for journalists.

Enter Virgil Griffith’s Wikiganda project and device WikiScanner, nominated for a Knight-Batten award for its use by Wired.com.

The devices could help journalists to separate fact from fiction on the site – and throw up some news leads in the process, the California Institute of Technology student told Journalism.co.uk.

1)How does WikiScanner work?
WikiScanner’s core functionality is the listing of ‘anonymous’ edits [of Wikipedia listings] via real-world organizations.

When you make an edit to Wikipedia, you have two choices: first, you can register and leave your username; or you can edit anonymously. But, when you edit anonymously, it uses your IP address – a number which identifies what computer network you are from – in lieu of a username. Wikipedia does this for convenience to distinguish your anonymous edits from someone else’s anonymous edits.

In essence, WikiScanner combines two databases: the list of all IP adresses that have made edits to Wikipedia; and what IP addresses belong to which companies. So with WikiScanner you can type a company name, and it shows you what edits have come from IP addresses owned by that company.

2) You recently upgraded Wikiscanner – why?
Pretty much everyone agrees that transparency is good for Wikipedia.

WikiScanner went a long way towards this, but it had two major flaws: it was too easy to hide from by either registering an account or by editing from home; and secondly, it took too much effort to find the interesting stuff [such as this article using the tool by Wired.com].

WikiScanner 2.0 addresses both of these defects. It tries to discover what organization registered Wikipedia accounts are coming from, and it uses some intelligence to highlight the edits that are likely to be the salacious conflict-of-interest stuff that people love to find.

3)You’ve also been working on a project call Wikiganda – what does that entail?

We all know that there are real-world organizations with radically opposed views. Wikiganda is a personal attempt to discover whether these divergent views spill over into sustained edit wars on Wikipedia. I do not know the answer to this question, but I’d like to find out.

What you do is input two real-world organizations that are ideologically opposed to each other, and Wikiganda lists the edits to pages that both organizations have modified. If the two sides are continually contradicting each other, the user flags it, and the world gets to inspect the results for themselves.

4)What’s the purpose of the project?
I am demonstrating that to have reliable information online doesn’t mean we need to erect walls blocking anonymous contributions. Instead, we can do back-end analyses of the contributions to filter out the bad stuff.

Overall – especially for non-controversial topics – Wikipedia seems to work. For controversial topics, Wikipedia can be made more reliable through techniques like this one. As for related approaches, I think colored text [a project that highlights Wikipedia articles in different colours according to their trustworthiness] is an immensely promising direction for combating disinformation in Wikipedia.

5) Do tools such as Wikiganda and WikiScanner enhance Wikipedia and other open online information sources by making them more transparent, or do they undermine them?
WikiScanner and Wikiganda do not undermine open information sources at all – it merely tells us the truth – that interested people and parties attempt to shape and influence them. From there we can take steps accordingly to address the problem.