Abstract: Implicit relevance is existed between multi-source geospatial vector data. However, the association relation often exists hidden, which is difficult to visualize directly as well as interactive presentation with spatial data, let alone query and analysis to obtain the required information. In view of this situation, multi-source geospatial vector data and statistical data is taken as research object. Firstly, the concept and classification of multi-source geospatial vector data association are defined. And then based on this, the multi-source geospatial vector data association model is designed, which can be divided into 3 sub-models:spatial association sub-model based on adaptive quadtree coding, spatial association sub-model based on geometric matching and spatial association sub-model based on semantic matching. The proposal of this model defines the association method between multi-source geospatial vector data, and lays theoretical foundation for the construction of association relations.

Key words: multi-source geospatial vector data, association concept, association model, adaptive quadtree coding, geometric matching, semantic matching