Interpretation of Spatial Data Bases using Machine Learning Techniques

摘要

The automation of spatial analysis procedures in a Geographic Information System (GIS) can be seen as a database interpretation problem. There is a great variety of possible applications for database interpretation, like re-interpretation of already existing data sets, extraction of implicit information, data conflation and data aggregation. The key problem in interpretation is the provision of the necessary knowledge in terms of object or process models. Usually it is given a priori and hard-coded in a system. In order to allow for a certain flexibility, generic models cna be applied. The use of Machine Learning gives the possibility of adaptation to new or changing situations. Such techniques can be utilized hwen implicit knowledge is given. In this approach, digital data sets are used as implicit knowledge sources.