Title :
TransGraph: Hierarchical Exploration of Transition Relationships in Time-Varying Volumetric Data
Author :
Gu, Yi ; Wang, Chaoli
Author_Institution :
Dept. of Comput. Sci., Michigan Technol. Univ., Houghton, MA, USA
Abstract :
A fundamental challenge for time-varying volume data analysis and visualization is the lack of capability to observe and track data change or evolution in an occlusion-free, controllable, and adaptive fashion. In this paper, we propose to organize a timevarying data set into a hierarchy of states. By deriving transition probabilities among states, we construct a global map that captures the essential transition relationships in the time-varying data. We introduce the TransGraph, a graph-based representation to visualize hierarchical state transition relationships. The TransGraph not only provides a visual mapping that abstracts data evolution over time in different levels of detail, but also serves as a navigation tool that guides data exploration and tracking. The user interacts with the TransGraph and makes connection to the volumetric data through brushing and linking. A set of intuitive queries is provided to enable knowledge extraction from time-varying data. We test our approach with time-varying data sets of different characteristics and the results show that the TransGraph can effectively augment our ability in understanding time-varying data.
Keywords :
data analysis; data mining; data structures; data visualisation; graph theory; probability; TransGraph; data visualization; graph based representation; hierarchical state transition relationships; knowledge extraction; navigation tool; occlusion free; time-varying volumetric data analysis; transition probability; visual mapping; Data mining; Data visualization; Hierarchical systems; Histograms; User interfaces; Time-varying data visualization; hierarchical representation; states; transition relationship; user interface.;
Journal_Title :
Visualization and Computer Graphics, IEEE Transactions on
DOI :
10.1109/TVCG.2011.246