Semantic data and geo-tagging the real world
The Japanese startup Tonchidot made a well-blogged ‘vaporware debut’ at TechCrunch50 last year. At the event they didn’t show a working app but a very feasible idea of an iPhone application - Sekai Camera -that allows people to tag the real world using the phone camera viewfinder as interface. It seems that they have finally unveiled it (or at least showed a working demo) at a fashion event in Tokyo this week.
This is a very cool and innovative idea and I bet we’ll see lots of startups trying to implement it in the next couple of years (just watch iPhone and Android app stores). In some sense there will be lots of tools for tagging (as there are already many geo-tagging tools like Flickr, Yellowpin, Panoramio, Brightkite, etc). However the real challenge is not making real world taggable but in how location-based search tools will extract the semantic data and make it relevant to what a someone needs in a specific context. The potential for noise, spam or just irrelevant annotations is imense. Someone like ought to deliver the right tags at the right moment. I remember Omnisio, a video tagging tool that was acquired by Google, and how much garbage they allowed users to tag - so much that the videos were just unwatchable.
