A Spatio-Temporal Knowledge-Discovery Platform for Earth-Science Data

The GPlates Geographic Information System (GIS) is a well-established tool for visualising and interacting with multimodal geo-data reconstructed through space and geological time. It overcomes the complexity of changing spatial reference frames due to plate-tectonic processes. Combining vast datasets in this manner is increasing the analysis complexity, with traditional visualisation-based approaches becoming ineffective in extracting key information and discovering hidden associations. This paper discusses the nature of these complexities, followed by the presentation of an extension to GPlates, involving the addition of interactive quantitative data-mining tools and capabilities, better suited to coping with the inherent analysis complexities. A case-study is used to demonstrate the system´s unique capabilities. The integrated software infrastructure has manifested itself as a powerful knowledge discovery platform, which has the potential to lead to new high-impact discoveries in the Earth Sciences.