I have been thinking around the concept for the thesis and the more I think about it the more I am excited by the idea of building this spatialised messaging system on mobile phones.

The core idea is to build a system where people can post messages attaching these messages to specific places where they happen to pass by. This was already implemented by GeoNotes but without any interesting long term study. Here I want to stress this concept further.

We can relate the context production to the communication intent and therefore at the social construction of meaning as last extent. The idea is to build something useful for people to play with, useful for me to do research and maybe cool for a company to put freely on their phones.

I see a great deal of convergence in this approach.

Now some notes on the things I see we can do with this phone system.

THINGS WE KNOW (Context):
– time of MSG/content
– position of content (x,y)
– position of the emitter, receiver
– words of the MSG
– history of the emitter, receiver (previous, current and future position and activities)
– self inputed categories
– friend of a friend chain
– social editing/rating of the message
– shape of the buildings and other geographical objects on the way

THINGS WE MAY INFER (statistically):
– average usage of that location
– clustering of the location (differences between places and locations)
– common interests based on common places frequentation

THINGS WE WANT TO KNOW (content):
– communication intent
– content categorization
– semantic description

In my idea of the system we have an initial ontology which provides the understanding of the content when the system starts (bootstrap), then over time messages gets categorised and put in relationship and new rules of semantic description are found by an agent (semantic network), this of course imply a better system of categorization of the messages which is therefore refined over time.

People using this system may also explicitly participate in the democratic definition of this ontology/categorisation, having the ability of voting and rating the accuracy of the system. From another point of view we could reason on this user ability to reflect upon the way they use the space if we provide this output as a community mirror. In addition we can have an evaluation of the effectiveness of people solving specific tasks related to space and group interaction.

These different approach are completely fine with me but we need to chose one of them so to circumscribe the approach and the field of interaction of the thesis.