Visualization and Visual Analytics


The above image is by Arek Socha from Pixabay.

There are currently seven teams of volunteers specialised in visualization and visual analytics: one generic support team, one analytical support team, four modelling support teams, and one disseminative visualization team. A report on the work by these teams between June and November 2020 can be found at arXiv:

The generic support team, which currently has nine members, focuses on developing a visualization server to provide visualization dashboards and plots to all modelling scientists and epidemiologists in the SCRC. Within the first month of its volunteering effort, it designed the overall software architecture and prototyped an experimental server with two dashboards and some 30 visualization plots. In the following month, it designed and prototyped several core components including an ontological database for storing the relations between visualization functions and data streams, a user interface with personal portals and search capability, and software agents for performing various housekeeping functions (e.g., automated generation of derived data, propagation of visualization functions to different data streams ).

The analytical support team, which currently has three members, identifies the analytical needs in different modelling workflows and decision workflows for combating COVID-19, develops and tests various data mining algorithms that may meet such needs, and then passes the effective algorithms to the generic support team for deployment. In anticipation that many data mining activities would be organised around time series, within the first eight weeks, the team implemented about ten different similarity metrics for time series analysis. Together with the generic support team, they developed several distance-based visualization techniques to accompany the similar metrics. When these were applied to the time series data associated with the 14 health boards in Scotland, the similarity and difference among the 14 regions became easily observable.

There are currently four modelling support teams, which have a total of 12 members. The four teams have established collaboration with different modelling scientists in the SCRC. Some teams have designed and prototyped model-specific visualization techniques, while other teams have completed their requirement analysis and have entered the phase of design and prototyping.

The disseminative visualization team currently has three members. They have surveyed the existing provision of visualization in the COVID-19 context, and concluded that more focused and engaging storytelling visualization can complement the existing provision by the governmental bodies and public services.