1、1,This work partially funded by NSF Grants IIS-9732897, IRIS-9729878 and IIS-0119276,Matthew O. Ward, Elke A. Rundensteiner, Jing Yang, Punit Doshi, Geraldine Rosario, Allen R. Martin, Ying-Huey Fua, Daniel Stroe,http:/davis.wpi.edu/xmdv,XmdvTool Interactive Visual Data Exploration System for High-d
2、imensional Data Sets,Worcester Polytechnic Institute,2,XmdvTool Features,Hierarchical visualization and interaction tools for exploring very large high-dimensional data sets to discover patterns, trends and outliers Applications: Bioterrorism Detection Bioinformatics and Drug Discovery Space Science
3、 Geology and Geochemistry Systems Monitoring and Performance Evaluation Economics and Business Simulation Design and Analysis Multi-platform support (Unix, Linux, Windows) Public domain software: http:/davis.wpi.edu/xmdv,3,Scale-up to High Dimensions: Visual Hierarchical Dimension Reduction Scale-up
4、 to Large Data Sets: Interactive Hierarchical Displays, Database Backend with Minmax Encoding, Semantic Caching and Adaptive Prefetching Interlinked Multi-Displays: Parallel Coordinates, Glyphs, Scatterplot Matrices, Dimensional Stacking Visual Interaction Tools: N-Dimensional Brushes, Structure-Bas
5、ed Brushing, InterRing,Xmdv: Main Features,4,Scale-Up for Large Number of Dimensions,Solution to High Dimensional Datasets: Group Similar Dimensions into Dimension Hierarchy Navigate Dimension Hierarchy by InterRing Form Lower Dimensional Spaces by Dimension Clusters Convey Dimension Cluster Informa
6、tion by Dissimilarity Display,5,Visual Hierarchical Dimension Reduction Process,6,A 42-dimensional Data Set,Dimension Hierarchy Interaction Tool: InterRing,A 4-Dimensional Subspace,Visual Hierarchical Dimension Reduction Process,7,InterRing - Dimension Hierarchy Navigation and Manipulation,Roll-up/D
7、rill-down Rotate Zoom in/out,Distort,Modify,8,Dissimilarity Display,Three Axes Method,Mean-Band Method,Diagonal Plot Method,Axis Width Method,9,Scale-up for Large Number of Records,Solution to Large Scale Datasets: Group Similar Records into Data Hierarchy Navigate Data Hierarchy by Structure-Based
8、Brushing Represent Data Clusters by Mean-Band Method Provide Database Backend Support using MinMax Tree, Caching, Prefetching,10,Interactive Hierarchical Display,Hierarchical Clustering,Structure-Based Brushing,11,Flat Display,Hierarchical Display,Interactive Hierarchical Display,Mean-Band Method in
9、 Parallel Coordinates,12,Flat Display,Hierarchical Display,Mean-Band Method in Parallel Coordinates,Interactive Hierarchical Display,13,Scalability of Data Access,Approach Attach database system to visualization front-end MinMax hierarchy encoding Key idea: avoid recursive processing Pre-computed Ca
10、ching Key idea: reduce response time and network traffic Prefetching Key idea: use application hints and predict user patterns Performed during idle time,14,Pre-compute object positions level-of-detail (L) extent values (x,y) preserve tree structure,New query semantics objects are now rectangles sel
11、ect objects that touch L select objects that touch (x, y) structure-based brush = intersection of two selections,Scalability of Data Access: MinMax Hierarchy Encoding,15,Purpose reduce response time and network traffic Issues visual query cannot directly translate into object IDshigh-level cache spe
12、cification to avoid complete scans Semantic caching queries are cached rather than objects minimize cost of cache lookup dynamically adapt cached queries to patterns of queries,Scalability of Data Access: Caching,16,Strategy Speculative (no specific hints) navigation remains local both user and data
13、 set influence exploration Adaptive (strategy changes over time) Evolves as more knowledge becomes available Non-pure (interruptible prefetching) leave buffer in consistent state Requirements non-pure prefetching + large transactions & small object size + semantic caching small granularity (object l
14、evel) speculative, non-pure prefetcher cache replacement policy + guessing method,Scalability of Data Access: Prefetching,17,Conclusions: Caching reduces response time by 80% Prefetching further reduces response time by 30% Designing better prefetching strategies might help further reduce response t
15、ime,Scalability of Data Access: Experimental Evaluation,18,Random Strategy,Direction Strategy,Focus Strategy,Mean Strategy,Exponential Weight Average Strategy,Vector Strategies,Data Set Driven Strategy,Localized Speculative Strategies,Scalability of Data Access: Prefetching,19,Xmdv System Implementa
16、tion,Tools C/C+ TCL/TK OpenGL Oracle 8i Pro*C,20,Publications (available at http:/davis.wpi.edu/xmdv),Jing Yang, Matthew O. Ward and Elke A. Rundensteiner, “InterRing: An Interactive Tool for Visually Navigating and Manipulating Hierarchical Structures“, InfoVis 2002, to appear Punit R. Doshi, Elke
17、A. Rundensteiner, Matthew O. Ward and Daniel Stroe, “Prefetching For Visual Data Exploration.”Technical Report #: WPI-CS-TR-02-07, 2002 Jing Yang, Matthew O. Ward and Elke A. Rundensteiner, “Interactive Hierarchical Displays: A General Framework for Visualization and Exploration of Large Multivariat
18、e Data Sets”, Computers and Graphics Journal, 2002, to appear Daniel Stroe, Elke A. Rundensteiner and Matthew O. Ward, “Scalable Visual Hierarchy Exploration”, Database and Expert Systems Applications, pages 784-793, Sept. 2000 Ying-Huey Fua, Matthew O. Ward and Elke A. Rundensteiner, “Hierarchical Parallel Coordinates for Exploration of LargeDatasets”, IEEE Proc. of Visualization, pages 43-50, Oct. 1999 Ying-Huey Fua, Matthew O. Ward and Elke A. Rundensteiner, “Navigating Hierarchies with Structure-Based Brushes”, IEEE Proceedings of Visualization, pages 43-50, Oct. 1999,
