Skip to Main Content
Advances in geographical information systems (GIS) and supporting data collection technology has resulted in the rapid collection of a huge amount of spatial data. However, known data mining techniques are unable to fully extract knowledge from high dimensional data in large spatial databases, while data analysis in typical knowledge discovery software is limited to non-spatial data. Therefore, the aim of the software system for spatial data analysis and modeling (SDAM) presented in this article was to provide flexible machine learning tools for supporting an interactive knowledge discovery process in large centralized or distributed spatial databases. SDAM offers an integrated tool for rapid software development for data analysis professionals as well as systematic experimentation by spatial domain experts without prior training in machine learning or statistics. When the data are physically dispersed over multiple geographic locations, the SDAM system allows data analysis and modeling operations to be conducted at distributed sites by exchanging control and knowledge rather than raw data through slow network connections.