Integration and evaluation of map generalization

Current generalization techniques are not applied in isolation in a manual context, and their integration remains an impediment to fully automated mapping. The graphical context of map items requires understanding in order to optimally integrate the various generalization operations. There is, in addition, a need to attend to issues of topogeneralization, or relevance of proximal features, in generalizing a map. Graphical tools to illustrate methods for blending and integrating of generalization techniques have been suggested. Optimal map design (i.e. the successful selection of generalization techniques which includes the selection of salient and contextual data) will depend on a clear understanding of complex spatial relationships as well as an understanding of how map images are interpreted. Data representations must reflect the principles of effective visualization and "until visual character can be analytically identified, it will remain a major stumbling block in automating generalization tasks'. -from Author