Y. H. Fua, M. O. Ward, and E. A. Rundensteiner. Hierarchical parallel coordinates for exploration of large datasets. In IEEE Visualization, pages 43--50, 1999.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....for information visualization and for visual data mining. General extensions to parallel coordinates during the long time that they have been an integral part of InfoViz research, several extensions to the original parallel coordinates have been presented. A hierarchical extension was proposed [5], for example, integrating automatic clustering to deal with really big data sets. Furthermore, a number of 3D extensions to parallel coordinates were presented [21] including extruded parallel coordinates or linking with wings, for example. Also, higher order parallel coordinates were proposed ....

Y.-H. Fua, M. O. Ward, and E. A. Rundensteiner. Hierarchical parallel coordinates for exploration of large datasets. In Proc. of IEEE Visualization '99, pages 43--50.

....in discontinuous bunches with meaningful qualities, the hierarchy imposed on it makes simple the selection of similar groups of objects within a dimension. Recent extensions of the parallel co ordinates technique such as the hierarchical parallel co ordinates metaphor developed by Fua et al. [63] can achieve the same effect. Another problem when using a dimension with a continuous scale can be seen in Figure 7.4. In the visualisation, leaf nodes and categories are ordered numerically for integer data sets, and alphabetically otherwise (so 12 doesn t come after 111 but before 132. Thus, ....

Fua, Y.-H., Ward, M. O. and Rundensteiner, E. A. (1999). Hierarchical Parallel Coordinates for Exploration of Large Datasets. Proc. of IEEE Visualization '99 (October 24-29, San Francisco, California, USA), IEEE Computer Society Press, 43-50.

....on the left hand side. Each vertical axis represents the topics of one time step. Each connecting line represents a sequential pattern. Colors are used to show the support of the patterns as well as sub patterns. The appearance of our visualization is somewhat similar to parallel coordinates [5][7] The horizontal axis (time) has a natural ordering, but the ordering of the elements on the vertical axis (topics) is essentially arbitrary. This is different from the situation in parallel coordinates where the ordering of the horizontal parallel axes is arbitrary, and the ordering of the ....

Ying-Huey Fua, Matthew O. Ward, and Elke A. Rundensteiner. Hierarchical Parallel Coordinates for Exploration of Large Datasets. In David Ebert, Markus Gross, and Bernd Hamann, editors, Proceedings IEEE Visualization '99, pages 43 -- 50, New York, NY, Oct 24 -- Oct 29, 1999. ACM Press.

No context found.

Y. Fua, M. Ward, and E. Rundensteiner, Hierarchical parallel coordinates for exploration of large datasets," Proc. Visualization '99, pp. 43-50, 1999.

No context found.

Y. Fua, M.O. Ward, and E.A. Rundensteiner. Hierarchical parallel coordinates for exploration of large datasets. Proc. IEEE Visualization, pages 43--50, Oct. 1999.

No context found.

Y. Fua, M.O. Ward, and E.A. Rundensteiner. Hierarchical parallel coordinates for exploration of large datasets. Proc. IEEE Visualization, pages 43--50, Oct. 1999.

No context found.

Y. Fua, M. Ward, and E. Rundensteiner. Hierarchical parallel coordinates for exploration of large datasets. IEEE Proc. of Visualization, pp.43-50, 1999.

....of data that can be handled by the tool. The limited presentation of data semantics within the visualization. Even though visualization systems have tried to tackle the problem of displaying large datasets [2] for example, pixel oriented displays [18] and hierarchical parallel coordinates [9]) effectively, the other issues have been given less attention. This paper tries to address the last two issues in an existing visualization system: Most of the current software systems for graphically presenting data are not built to handle a wide range of data types and hence they tend to be ....

Y. Fua, M. Ward, and E. Rundensteiner. Hierarchical parallel coordinates for exploration of large datasets. In Proc. of Visualization '99, p. 43-50, San Francisco, California, USA, Oct. 1999.

....The clutter problem not only exists in visualizing data sets with high dimensionality, but also when visualizing data sets with a large number of data items. Our previous work has addressed the clutter problem in the latter situation using an Interactive Hierarchical Display (IHD) framework [7, 8, 24]. The underlying principle of this framework is to develop a multi resolution view of the data via hierarchical clustering, and to design extensions of traditional multivariate visualization techniques to convey aggregation information about the resulting clusters. Users can then explore their ....

....hierarchical clustering, and to design extensions of traditional multivariate visualization techniques to convey aggregation information about the resulting clusters. Users can then explore their desired focus regions at different levels of detail, using our suite of navigation and filtering tools [7, 8]. Inspired by the IHD framework, we now propose a new methodology for dimensionality reduction that combines au Figure 2: A subset of the Census Income data set (42 dimensions, 200 data items) in Parallel Coordinates. Individual data items cannot be seen clearly. tomation and user interaction ....

[Article contains additional citation context not shown here]

Y. Fua, M. Ward, and E. Rundensteiner. Hierarchical parallel coordinates for exploration of large datasets. Proc. of Visualization '99, p. 43-50, Oct. 1999.

....semantic caching [24, 8] for the advantages it o ers over the traditional caching systems. To further improve the performance, we propose to exploit characteristics of visualization environments to prefetch the data for the visualization tools. We have incorporated these features into XmdvTool [47, 13, 14, 50], a freeware visual tool for multivariate exploration. We also compare an array of di erent prefetching strategies to determine their relative e ectiveness for both synthetic user traces and real users of our system. Our results show that signi cant improvement can be achieved for visualization ....

....of queries formulated via a visual query tool in order to optimize the contents of the 1 cache. We support features necessary for visualization, by making our prefetching solutions speculative and non pure. The proposed caching and prefetching techniques have been incorporated into XmdvTool [47, 13, 14, 50], a freeware visual tool for multivariate exploration. We have also ran experiments to evaluate the performance of the prefetching strategies both with various synthetic user traces as well as with real users of our system. Results show that the proposed strategies indeed improve the performance ....

[Article contains additional citation context not shown here]

Y. H. Fua, M. O. Ward, and E. A. Rundensteiner. Hierarchical parallel coordinates for exploration of large datasets. IEEE Proc. of Visualization, pages 43-50, Oct. 1999.

....postulates that any method that displays a single entity per data point invariably results in overlapped elements and a convoluted display that is not suited for the visualization of large datasets. A new approach has been proposed recently for displaying and visual exploring large datasets [6]. The idea is to present data at different levels of detail based on hierarchical clustering the initial datapoints. The problem of cluttering at the interface level is solved by displaying one level of detail only at a time. However, such hierarchical summarizations result in increasing the size ....

....a software package (http: davis.wpi.edu xmdv) designed for the exploration of multivariate data. The tool provides four distinct visualization techniques (scatterplot matrices, parallel coordinates, glyphs and dimensional stacking) with interactive selections and linked views. Our recent efforts [6, 7] have produced versions of these techniques that allows multi resolution data presentation. 2.1 Visual Brush Based Exploration Selection is a process whereby a subset of entities on a display is isolated for further manipulation, such as highlighting, deleting, or analysis [20] Brushing is the ....

Y. H. Fua, M. O. Ward, and E. A. Rundensteiner. Hierarchical parallel coordinates for exploration of large datasets. IEEE Visualization, pages 43--50, Oct. 1999.

....As a generalization, 10] postulates that any method that displays a single entity per data point invariably results in overlapped elements and a convoluted display that is not suited for the visualization of large datasets. A new approach has been proposed recently for displaying large datasets [9]. The idea here is to present data at different levels of detail based on applying an aggregation function to a hierarchical structure that results from a proximity clustering process. The problem of cluttering at the interface level is solved by displaying only a limited set of aggregates at a ....

....approximate shape formed by chaining the leaf nodes. The colored bold contour across the tree delineates the tree cut that represents the cluster partition corresponding to the specified level ofdetail. XmdvTool uses a proximity based coloring scheme in assigning colors to the partition nodes [9]. In this scheme, a linear order is imposed on the data clusters gathered for display at a given level of detail. This linear order is directly derived from the order in which the tree is traversed when gathering the relevant nodes for a given level of detail. Colors are then assigned to each ....

[Article contains additional citation context not shown here]

Y. H. Fua, M. O. Ward, and E. A. Rundensteiner. Hierarchical parallel coordinates for exploration of large datasets. IEEE Proc. of Visualization, pages 43--50, Oct. 1999.

....data. Keywords: Hierarchy Encoding, Hierarchical Drill Down and Roll Up, Recursive Query Processing, Relational Databases. 1 Introduction Recently, we have introduced a new technique for visual exploration and analysis of large datasets based on a hierarchical structuring of the information [FWR99a] The technique, incorporated in This work is supported under NSF grant IIS 9732897. 1 the XmdvTool visualization system [War94] is part of our on going efforts in the area of scaling visualization technology to be effective for large datasets. The work presented in this paper has ....

....used only the main memory to store the necessary information during exploration. However, the assumption that all of this information can fit in the main memory 6 is not valid for large data sets. To overcome the visualization limitation we proposed a hierarchical model for exploring the data [FWR99a, FWR99b] The data points are recursively grouped into clusters based on a proximity function. The clustering process generates a hierarchical structure in which each hierarchy level corresponds to a level of abstraction at which the data can be visualized. Each cluster maintains some statistical ....

Y. H. Fua, M. O. Ward, and E. A. Rundensteiner. Hierarchical parallel coordinates for exploration of large datasets. IEEE Proc. of Visualization, pages 58--64, October 1999.

....brushes and the containment criteria. Section 4 describes the creation and manipulation of our brush. Following that in Section 5, we illustrate the usefulness and generality of our tool by applying it to two hierarchical visualization techniques, namely, hierarchical parallel coordinates [2] and tree maps [7, 12] We conclude by summarizing our contributions and outlining plans for future work. 2 Brush Basics Selection is a process whereby a subset of entities on a display are isolated for further operations, such as highlighting, deleting, or analysis. Wills [15] defined a taxonomy ....

....approximate shape formed by chaining the leaf nodes. The colored bold contour (see (b) across the tree delineates the tree cut that represents the cluster partition corresponding to the specified level of detail. We use a proximity based coloring scheme in assigning colors to the partition nodes [2]. In this scheme, a linear order is imposed on the data clusters gathered for display at a given level of detail. This linear order is directly derived from the order in which the tree is traversed when gathering the relevant nodes for a given level of detail. In our implementation, we adopt the ....

[Article contains additional citation context not shown here]

Y. Fua, M. Ward, and E. Rundensteiner. Hierarchical parallel coordinates for exploration of large datasets. Proc. of Visualization '99, Oct. 1999.

No context found.

Y. H. Fua, M. O. Ward, and E. A. Rundensteiner. Hierarchical parallel coordinates for exploration of large datasets. In IEEE Visualization, pages 43--50, 1999.

No context found.

Y.-H. Fua, M. O. Ward, and E. A. Rundensteiner. Hierarchical parallel coordinates for exploration of large datasets. Technical report, Worcester Polytechnic Institute, 1999.

No context found.

Y. H. Fua, M. O. Ward, and E. A. Rundensteiner. Hierarchical parallel coordinates for exploration of large datasets. In IEEE Visualization, pages 43--50, 1999.

No context found.

Ying-Huey Fua, Matthew O. Ward, and Elke A. Rundensteiner. Hierarchical parallel coordinates for exploration of large datasets. In David Ebert, Markus Gross, and Bernd Hamann, editors, IEEE Visualization '99, pages 43--50, San Francisco, 1999. IEEE.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC