Do you think was ever touched by a real person? And how many will ever be seen by a single person?

November 8th, 2009 by Laura Leave a reply »

I read McNiche: On the perils of scaling down a mass model a day or two ago and it has been bothering me ever since.

There was an observation or two about WikiCity that made me really think about Fan History and some of our automated content creation:

How many of those millions of “Web site pages” do you think was ever touched by a real person? And how many will ever be seen by a single person?

About 700,000 of Fan History’s 808,000 articles were created using automated methods that didn’t have a Fan History admin manually curating information.  We did this for FanFiction.Net people articles, for episode articles, for musician/band related articles.  I’d guess that of those 700,000 articles created by bots, roughly 50,000 have been touched by human hands.

Last month, there were 37,272 pages were viewed a total of 176,272 times.   That 37,272 works out to roughly 4% of our pages being accessed in October 2009.  Year to date, 183,755 pages were viewed a total of 1,786,830 times.  This works out to 22% of all our articles being accessed this year.  (Which basically supposes that only articles on the main name space were accessed.)  That doesn’t feel all that impressive.  We’d really, really like to improve on that.  It is one of our goals.  That and to get more people to edit pages that were created through automation.

The article talks about how scaling down these projects might be the best option because of all those dead pages that aren’t touched by human hands.  I’m just not sure for some large scale wiki projects that rely on automation like Fan History or AboutUs would benefit from this sort of thinking.  One of the things we’ve found on Fan History is that by having a large number of consistently formatted articles about certain topics is that we’ve been able to increase our number of contributors and have been able to get people to edit those articles.  It also helps with modeling; new contributors are able to look at those articles and create ones of their own.  We see this happen on a daily basis on Fan History: People discover articles about themselves or groups they are involved with and improve an article.  The article gets curated.  After that happens, the person doing it frequently links to it, tells others about it and then their friends go and create articles about their interests OR edit existing articles.  With out that automation, that wouldn’t happen.  If we were better at self promoting, our particular situation would improve and we’d have more people editing and curating and reporting and documenting.  The problem isn’t the automation; the problem is our ability to get others involved.

blog comments powered by Disqus

Canonical URL by SEO No Duplicate WordPress Plugin