CATEGORICAL DATABASE INTEGRATION AND GENERALIZATION BASED ON THEMATIC ATTRIBUTES

Y. Liu, L. Jiao

Wuhan Univ., GIS, Wuhan, PRC.

lmjiao027@163.com

Integration and generalization of thematic database are currently hot topics in database application.

At first, this paper discusses the methodology and process of integrating local (cities and counties) thematic databases established separately into one integrated database with seamless connection, semantic identity and scientific classification.

Secondly, it puts forward supporting data model and analysis model for semantic neartude. This paper uses a hierarchic semantic similarity model to evaluate semantic neartude.

And then it gives database generalizing rules and restricted factor system, as well as operators and process. Constraints in categorical database generalization include geo-spatial model constraints, objects constraints and relation constraints. We create corresponding database transformation units based on deferent constraints, then construct operations and process based on these transformation units.

Finally, it takes 86 local databases of agriculture land classification in Hubei province as an example to implement database integration and generalization from 1:50000 to 1:500000.