Application of Visualization Cube to Analysis of Interaction Pattern in Exploratory Data Analysis of Spatiotemporal Trend Information

摘要

his paper analyzes interaction pattern when a user performs exploratory data analysis of spatiotemporal trend information. Recent growth of the Web has brought us various type of information, among which trend information is one of the important information, because it is crucial for our decision making. In order to support users who want to make use of such information, an interactive information visualization system is useful for supporting their exploratory data analysis. However, it is difficult to understand user's behavior when performing exploratory analysis, even though it is important information for improving the support capability of the system. In order to solve these problems, the concept of visualization cube has been proposed as a basis for establishing adequate interaction model of exploratory data analysis of spatiotemporal trend information. This paper focuses on the usability of the visualization cube in terms of analyzing users' interaction pattern when performing exploratory data analysis of spatiotemporal trend information. The prototype system is developed based on the visualization cube, which was used by 5th-grade elementary school children and graduate/ undergraduate students. The frequent interaction patterns are extracted from the interaction logs. The result shows the visualization cube is useful as a basis for analyzing users' interaction pattern.