[2015-2020] I was the sole researcher resourced on the EPSRC funded Automated Map Content Selection project. Over the course of my PhD I conducted research with 64 participants over the course of 7 complex navigation studies. I developed a conceptual framework for automatically generating highly contextual navigational information based on traveller context, and formalised this framework into a prototype system for generating interactive multi-scale maps.

My approach centred on modelling spatial features as variables in causal graphs, and using statistical methods for then predicting, given the current journey context, the probability that a feature should be in a given map view. This is visualised below, where a given spatial extent at a given scale can be understood in terms of a lateral slice across this probabilistic view space. The research can be found here.