Location Based Services Confront The Empty Room Problem
By Mike Masnick, Tue Jun 14 01:00:00 GMT 2005

Now that more and more handsets can make use of location-based data, someone needs to populate those services with data. Some are taking unique approaches to filling databases with data.

There are all sorts of intriguing attempts at creating location-based services with the idea that people can create their own data for others to use. While the idea is intriguing, it does face some challenges. The biggest, of course, is the "empty room" problem. If there is no data in the system, no one's going to step up and add any data as it doesn't seem worthwhile. No one will use it. It's the catch-22 of any system that requires lots of users to input data. If there isn't an initial data set it's hard to jumpstart the effort.

In some cases, as with basic driving directions, the data comes from companies that have spent a lot of effort building up the database themselves. However, that doesn't work for more niche type applications and often requires substantial upfront expense by the company with the data -- meaning it doesn't come cheaply, either.

While it may not be interesting to everyone, some in-line skaters have realized that they could use location-based info to their advantage. The problem is that there just isn't very good info on places to skate -- so they've created a system where in-line skaters can easily and automatically populate a database of skating trails based on a GPS system and mobile phone. The skater just needs to have the system on them while they skate, and he or she can upload the trail -- and later add additional notes if so desired.

What's happening here is that the creation of the initial database is being automated to some extent. By making it easy to record and post the data, the project takes away many of the hurdles that make the empty room problem such a big issue. This way, the data gets entered quickly and easily -- and the incentive is for the skaters to add additional personalized data to the automated data. Of course, such a system doesn't necessarily need to be limited to skaters, but could just as easily apply to walkers, runners and bikers. The trick is simply lowering the barrier to creating that data and allowing the data to build up in as automated a way as possible.

While it may seem like a very niche application, it's these small niche applications that will create the data that makes real location based services useful in the future. Lowering the barriers to entering data by including some portion of automated data creation is an important step forward.