Last week I was in TGV, going to INRIA in the south of France, and wanted to post a tweet. Since my iPhone application offers to publish my current location along with the tweet it occurred to me how nice would it be to actually disclose my current location. To say to all my friends that I was traveling to Sophia-Antipolis (near Nice). Of course, the iPhone application is not capable to capture this kind of information. It can only capture my current location, in terms of latitude and longitude. But in this case, the exact geographical point is totally irrelevant – what I need to state is that I am in a train with a certain direction, not that at I am passing through a particular spot at a particular moment.
I do not know of an application that could serve for such a use case. Neither do I know of a model that could capture such semantics. In the current version of the Ontology of Online Presence, we have used a simple geographical spot to describe user’s current location. The scenario that I have presented here obviously makes this insufficient, so I will work on a way to improve on a model and allow statement about being on the move.
The idea is interesting and several aspects have already been treated in the Bell Labs. However the main difficulty here is not the ontological representation of the concept, but in the algorithm that can automatically infer this kind of high level information from raw data. More concretely, a frequent change of cellular ID-s could mean that the user is moving. Direction can be learned from GPS data. Also, a calendar can enrich, validate the resulted user situation, that can automatically “tag” the user’s status message.
Link | September 8th, 2009 at 2:01 pm
Certainly, many ways can be thought of to infer such a thing (train ticket information along with what you have mentioned).
Are there any papers (from Bell Labs) where I could read more about the aspects you have treated?
Link | October 2nd, 2009 at 10:31 pm