How do social networks differ when inferred from different sources? Do they each give equally valid social networks in different contexts or are some sources of data inherently incomplete. How can we compare and quantify these? Mobile phone data will be one of the inputs to this project
Does the movement of cows affect the spreading of tuberculosis (TB)? What is the effect of animal movements restriction policies on the movement of cattle and the spread of disease. This project will involve working with the UK Animal and Plant Health Agency (APHA) who have provided data.
Geographically-weighted statistics help us understand geographical phenomena, but these are rarely used in an interactive exploratory visualisation context. Is there a role for them in this context? What are the best interactive visualisation designs? How well do they support exploratory analysis?
How can existing visual analysis approaches be extended to facilitate the concurrent analysis of multiple non-physically-linked datasets? How can the semantics of the data and interpretations coming from the application domain be instrumental in enabling such analysis processes?
How can interactive visualisations serve as a medium to externalise and exchange observations and the analytical reasoning that led to these? Can such representations serve as a provenance / reproducibility mechanism within data-rich scientific disciplines?
Exploratory visualization has been shown to be highly effective at revealing structure, patterns and inconsistencies in new and complex data. In fact, early visual inspection of a data set is now de regueur for most applied research domains. But when it comes to the business end of most serious scientific endeavour — making evidence-based claims — researchers (even those within visualization!) tend to restrict themselves to a limited, and often misused set of techniques and reporting conventions. We’d welcome applications from motivated candidates wishing to evaluate the use of visualization for communicating scientific evidence. In particular, we’d like to know whether visualization can be used to support understanding of simple statistical issues: effect, variability and uncertainty.