Concepts for an Agent Based Visualization & Characterization Tool

Abstract

The overall goal of this research is to develop a tool for adapting visualizations based on the goals, performance of the user, and characterizing information and meta-information in a holistic manner to support decision making in a complex environment. Our system features such assets as state-of-the-art visualization formats, a unified architecture that incorporates the information and domain into the visualization decisions, and user interaction. The holistic analysis, visualization, & characterization assessment tool (HAVCAT) uses a holistic view of the information system in that not only can primary and secondary effects/relationships be found and displayed, but even tertiary, and possibly beyond. The intelligent agent structure that underlies the tool is able to search through, using semantic meanings, a myriad of possibilities that may benefit the user. The tool analyzes alternate configurations that the user may not even be aware are possible, propose visualizations which incorporate human interaction, characterize the data for adaptation purposes, and assess the effectiveness of each by testing against defined metrics