Are We Safeguarding Social Data?

Recently, the social bookmarking service Ma.gnolia crashed in a big way, losing about half a terabyte of data in the process, including the backup of that data.

Ma.gnolia founder Larry Halff admits that the backup system just wasn’t solid enough. Halff has also posted some tools and information on the main Ma.gnolia page that should help users recover some of their data. However, it sounds as though the future of the service may now be in question.

The episode raises some interesting questions. Sebastien Paquet, a professor of computer science at the Universitédu Québec à Montréal, posted nine ways to protect data stored in the cloud, inspired by his own loss of data stored using Ma.gnolia. He and others have taken the opportunity to reexamine their trust in Web 2.0 services.

I say that the problem goes much further. For example, how safe is the data behind something like Wikipedia? Ed Chi, a senior research scientist at the Palo Alto Research Center (PARC), and his colleagues have been studying Wikipedia since about 2006 (see my article on their tool, WikiDashboard). In the course of their work, they struggled to get access to live data from Wikipedia that could make their tool more up to date and useful.

Chi told me, “What was difficult about doing this is that the Wikipedia Foundation has a lot of work done by volunteers, and getting access to the live data required application to a special research server, which is only occasionally administered on a volunteer basis. It’s amazing that it works at all, as the server really doesn’t get much attention and needed maintenance. It is such a valuable resource, so it really should get more attention and caring and feeding.”

Chi also suggested that Wikipedia’s data is not getting backed up as often as it should be. Losing data from Ma.gnolia is a personal loss for many, but Wikipedia is one of the great achievements of the digital age, and losing even a day’s worth of edits would be significant. We can’t afford to lose sight of the basic upkeep that will allow us to hold on to what we create.

Losing a large volume of socially created data would be a cultural loss. Wikipedia, Flickr, Twitter, WordPress, etc. store troves of data that will be valuable to future generations. I hope the data is being treated accordingly.