Streamlined workflow for large-scale interactive geographic visual analytics

Abstract

Cartography and computer science powerfully attract each other to band together into a melting pot stimulating research in interdisciplinary fields such as Geographic Visual Analytics. Cartography with its rich history of practice and science brings along visualisation knowledge and large data sets. Computer science with its short but vigourous evolution brings along data structures, algorithms and hardware to make the visualisations and large data sets come alive for unprecedented user interaction. However, it turns out that the traditional workflow suffers from noticeable delays and read-only limitations which practically hinder the every-day convenient and fluent interaction with large data sets. We show how to streamline the traditional workflow by eliminating the intermediate data conversion step through switching to a native XML database. We suggest to use a RESTful interface providing scalable temporal read-write access. Finally, we provide preliminary measurements based on our prototype implementation named TreeTank providing both compressed storage and fast SVG delivery.