Abstract: As one of the basic aspects of spatial data quality, the maintenance of logical consistency plays an important role in fields such as spatial database construction and updating. By comparing with the topologic conflicts existed between homogeneous objects, the consistency problems between heterogeneous objects, such as spatial pattern mismatching, conflicts of semantic relationship, are difficult to be handled. From the view of Tobler’s first law of geography, this study aims to solve inconsistencies of spatial data by introducing the rules of spatial dependent distribution between differentgeographical entities. Firstly, we identify three different spatial dependency situations (i.e. location-sharing, structure-associating, and area-including)and further to obtain the constraint rules from those spatial knowledge.Then, we propose inconsistency detection and correction methods based on the three types of constraint rules. This study recognizes the quality of spatial data from geographic view and obtains the constraint rules using knowledge engineering theory, which are beneficial for extending classic theory ofspatial data quality.

Key words: data quality, heterogeneous objects, conflict detection, consistency correction, constraint rules