LoboVault Home

A conceptual framework for visual data mining, with continuous semantic zooming


Please use this identifier to cite or link to this item: http://hdl.handle.net/1928/10321

A conceptual framework for visual data mining, with continuous semantic zooming

Show full item record

Title: A conceptual framework for visual data mining, with continuous semantic zooming
Author: Xia, Shan
Advisor(s): Caudell, Thomas P.
Committee Member(s): Pattichis, Marios S.
Sen, Pradeep
Goldsmith, Timothy E.
Department: University of New Mexico. Dept. of Electrical and Computer Engineering
Subject(s): Visual data mining
continuous semantic zooming
LC Subject(s): Data mining.
Information visualization.
Semantic computing.
Human-computer interaction.
Degree Level: Doctoral
Abstract: With the growing pervasiveness of modern computer technology, sensors, and networks, the amount of data collected in and about society, medicine, and the environment, for example, is exploding. Large-scale data sets are typically complex making the interesting or novel information hidden within them hard to discover. Data mining algorithms and information visualization methodologies can help people explore these data sets by extracting knowledge and representing it in "meaningful" ways. The integration of data mining algorithms and information visualizations approaches is a growing field of research and application. In this research, we propose a conceptual framework to help in the understanding of unified data mining and information visualization systems. To explore this framework, we implement an archetypical hierarchical prototype that exposes the internal parameters of the framework allowing the quantitative evaluation of its usefulness. It is hypothesized that unified systems designed in the context of this conceptual framework will enhance human data exploration over traditional data mining systems. The result of a pilot study that uses an information visualization technique referred to as continuous semantic zooming in the visualization portion of the prototype is presented that begins to support this hypothesis. Moreover, this project conducts a larger scale human subject experiments to further assess this hypothesis, and empirically characterize the effects on human performance by unified systems designed in this manner. The main experiment study yields promising results and further supports the hypothesis.
Graduation Date: December 2009
URI: http://hdl.handle.net/1928/10321

Files in this item

Files Size Format View
Dissertation_SXia.pdf 25.19Mb PDF View/Open

This item appears in the following Collection(s)

Show full item record

UNM Libraries

Search LoboVault

Advanced Search


My Account