Spatial Concept Lattices: An Integration Method in Model Generalization

Abstract

The ability to view and analyze data of different detail and from different perspectives, and to move dynamically from one scale to another requires modeling geographic information at different generalization levels. On this account, in the framework of model generalization, the concept of multi-scale database is adopted to provide a consistent multiple representation of existent mono-scale representations. Spatial Concept Lattices are propounded as a new approach to thematic generalization through the semantic integration of multiple classification schemata and the creation of a multi-scale, multi-context database. The methodology presents in an explicit and systematic manner the integration of classification schemata, which exhibit differences in spatial and thematic resolution. In order to comprehend the stepwise SCL methodology, an actual example is used to demonstrate the integration of three independent land cover/land use classification schemata. The integration process is part of model generalization, since the resulting hierarchical integrated schema supports various levels of thematic resolution and represents geographic space from different application perspectives.