Wikipedia’s ascendancy to the top of a large pool of online reference sites has come to an end, new research shows. But perhaps even more alarming to eleventh-hour term paper writers and lazy journalists everywhere, the user generated encyclopedia may have started a slow decline, as growth has leveled and the nature of the community has made it less welcoming to new contributors.
“It’s easy to say that Wikipedia will always be here,” says Dr. Ed Chi, a senior scientist at the Palo Alto Research Center, told New Scientist. “This research shows that is not a given.”
Eight years after launching, Wikipedia contains nearly 3 million articles submitted by users themselves and edited by others in the Wikipedia community. But while the site is still wildly popular for those trying to figure out who played in Super Bowl XXII, the name of Pink Floyd’s original frontman, or how a convection oven works, the explosive growth that characterized its early years leveled in 2006 at around 60,000 new articles per month, declining by nearly a third since then. But the trends within the community itself are far more troublesome. The year after the amount of new content flattened, the number of edits per month plateaued as well at around 5.5 million. Meanwhile, the number of users making edits leveled off at around 750,000 monthly.
The data suggest the Wikipedia community is becoming resistant to new content and new editors. Those passive editors who make just a single change per month see around a quarter of their changes erased or modified by other, more active editors (the rate was just 10% in 2003). The resulting exclusion of more varied contributions shifts the balance of power on Wikipedia to those fewer active editors, and in turn could make Wikipedia more like a fraternity than a community-driven social encyclopedia. Over time, the Palo Alto team warns, the quality of Wikipedia could degrade as there would be fewer editors checking for errors and correcting the vandalism that accompanies user-gen Web sites.
The non-profit Wikimedia Foundation responsible for running Wikipedia has launched an internal review to make sense of the data, but says the numbers could be deceiving. For instance, the higher number of modifications to passive contributors’ edits could be the result of human editors removing promotional text added to articles by spam programs, an increasing problem on Wikipedia. In the meantime, the Palo Alto team will present its research at the International Symposium on Wikis and Open Collaboration this October, though Dr. Chi has already begun blogging about the team’s analysis.
[via New Scientist]