Quality-Aware Visual Exploration Tools

点击翻译按钮获取中文摘要

Visualization has been identified as a critical component in the process of interactive exploration and mining of complex data repositories. This project focuses on the issue of quality at all stages of the visual exploration process. Types of quality considered include quality of the data itself, including reliability, age and completeness; quality of the transformation of the data into aggregated data abstraction; quality of data querying process by controlling of quality-of-service of retrieved data; and quality of mapping of data attributes to graphical objects and their screen layout resulting in different perceptible structures and various degrees of clutter.To achieve a ground-breaking holistic approach towards handling quality across the visual exploration pipeline, this project develops at each stage: (1) Metrics and models for measuring and capturing quality; (2) Algorithms for quality optimization based on selected quality metrics; (3) Visualizations to convey quality information for both data and structure spaces; and (4) Interactive tools to enable the user to control the optimization processes and specify preferences when trade-offs exist.Intellectual merit involves the definition, calculation, optimization, and visual presentation of quality metrics and attributes of both the information content and the structure space at all stages of the visualization pipeline. The expected results will have an impact in the application domains that routinely perform exploratory data analysis. The resulting tools enable effective visual exploration by explicitly exposing and integrating the quality of the data, which, while often ignored in existing tools, can have a tremendous impact on the decision-making process. The results of the project will be disseminated via the Internet (http://davis.wpi.edu/~xmdv).