Scalable Visual Hierarchy Exploration

Abstract

More and more modern computer applications, from business decision support to scientific data analysis, utilize visualization techniques to support exploratory activities.Most visualization tools do not scale well with regard to the size of the dataset upon which they operate. Specifically, the level of cluttering on the screen is typically unacceptable and the performance is poor. To solve the problem of cluttering at the interface level, visualization tools have recently been extended to support hierarchical views of the data, with support for focusing and drilling-down using interactive brushes. To solve the scalability problem, this thesis investigates how best to couple such a visualization tool with a database management system without losing the real-time characteristic required in any interactive application.This integration must be done carefully, since visual user interactions implemented as main memory operations do not map directly into efficient database operations. In our context, the main efficiency issue is to avoid the recursive processing required for hierarchical data retrieval. For this problem, we have developed a tree labeling method, called MinMax tree, that allows the movement of the on-line recursive processing into an off-line precomputation step. Thus, at run time, the recursive processing operations translate into linear cost range queries. Secondly, we employ a main memory access strategy to support incremental loading of data into the main memory. To further reduce the response time in the system, we have designed a speculative non-pure prefetcher that brings data into memory when the system is idle. The techniques have been incorporated into XmdvTool, a multidimensional visual exploration tool, in order to achieve scalability.The tool efficiently scales up now to datasets of the order 10 5 10 7 records. Lastly, we report experimental results that illustrate the impact of the proposed techniques on the system's overall performance. AcknowledgementsMany thanks to my professors and to my friends for their support.

Prefetching for visual data exploration

“…The objects (data clusters) now have a spatial representation that makes them selectable by the active window. As shown in [23,24], this additional information, consisting of a level value …”

“…The newly added modules are written in C with Pro*C (embedded SQL) primitives for Oracle8i. First, an off-line process clusters the flat data sets and then transforms the hierarchical data into MinMax trees [24], a pre-coded indexing structure that allows us to express hierarchical navigation as range queries. The transformed data is then loaded into the database.…”

Multiple-Scale Visualization of Large Data Based on Hierarchical Clustering

“…In fact, the overall premise is that users have a deeper understanding about their data when they interact with the presented information and view it at different levels of abstraction [2]. During the last two decades, many interactive visualization techniques and system have been emerged [1], [3], [4]. As large data sets become more and more common, with the size over 1K, it has been clear that most of the current visualization approaches lose their effectiveness due to they have no ability to visualize and manage the large number of data points simultaneously.…”

A progressive vector map browser for the web

“…Several authors have also investigated suitable techniques for memory caching of spatial data with and without prefetching 31,32 , as well as methods appropriate for handling multi-resolution vector data 25 .…”