Skip to Main Content
It is a laborious process to quantify relationship patterns within a feature-rich archive. For example, understanding the degree of neuroanatomical similarity between the scanned subjects of a Magnetic Resonance Imaging (MRI) repository is a nontrivial task. In this work we present a Coordinated Multiple View (CMV) system for visually analyzing collections of feature-rich datasets. A query-based user interface operates on a feature-respective data scheme, and is geared towards domain experts that are non-specialists in informatics and analytics. We employ multi-dimensional scaling (MDS) to project feature surface representations into three-dimensions, where proximity in location is proportional to the feature similarity. Through query feedback and environment navigation, the user groups clusters that exhibit probable trends across feature and attribute. The system provides supervised classification methods for determining attribute classes within the user selected groups. Finally, using visual or analytical feature-wise exploration the user determines intra-group feature commonality.