Scalable Architecture for Visual Data Exploration.

摘要

Intelligence analysts in the areas of defense and homeland security are now faced with the difficult problem of discerning the relevant details amidst massive data stores. We propose a component-based visualization architecture that is built specifically to encourage the flexible exploration of geospatial event databases. The proposed system is designed to deploy on a variety of display layouts, from a single laptop screen to a multi-monitor tiled-display. By utilizing a combination of parallel coordinates, principal components plots and other data views, analysts may reduce the dimensionality of a dataset to their most salient features. Of particular value to our target applications are understanding correlations between data layers both within single views and across multiple views. Our proposed system aims to address the limited scalability associated with coordinated multiple views (CMVs) through the implementation of an efficient core application which is meant to be extended by the end-user rather than a programmer.