RELEVANCE OF VISUAL EXPLORATION FOR STRENGTHENING SPATIAL THINKING SPATIAL KNOWLEDGE EXPLORATION

摘要

Tremendous amount of spatial data reside in operational or legacy data stores of public and private institutions. These databases contain topographic maps, aerial photos, satellite images, medical data, laser/lidar scanner data, video images among others. In addition to spatially referenced data, there are links from spatial objects to non-spatial data such as census, economic, security, and statistical information. It is costly and often unrealistic for users to examine spatial data in detail and search for meaningful patterns or relationships among huge amount of data. Spatial data mining (SDM) aims to automate such a knowledge discovery process in large databases along with visual exploration techniques.Visual exploration allows user to interact with huge spatial datasets to recognize spatial distributions and relationships and to extract meaningful information by facilitating visualization of spatial patterns and processes, and other cognitive skills such as visual thinking, visual comparison. This paper introduces visual exploration methods complementing spatial data mining methods, visual representations of knowledge exploration on historical ship accident data at Istanbul Strait, crime data in Ankara and earth quake data of Turkey. These methods are used to visualize and explore spatio-temporal data, to support visual thinking and knowledge exploration, to reveal unknown relationships.