Research On Multi-Scale Data Model For Map Generalization Based On Semantic Network

With the development of the information society, the production mode of traditional map also turns to the direction of digitization, automation and intelligent. In this process of change, map generalization as the problem that cannot be avoided in the geographical information system and digital map production has never seemed more necessary. Continuous production of series scale map reflects the issues of multi-scale representation of spatial data and the requirements of GIS fast acquisition and update of map data. And map generalization is an ideal strategy to solve this problem. Map generalization is the process of the abstraction and simplification of objective things. In essence, it is the thinking process of human brain. Map generalization contains two types of information, one is the basic map information described their own; another is the information about generalization rules. To establish a new model for efficient storage of data of these two types of information, it is not only conducive to describe complex geographical entities, but also conducive to transform spatial scale, and is one of the hot research on map automatic generalization at present.This paper focuses on formal expression of map generalization and establishing a multi-scale data model. To achieve this goal, author carries out a bit of work:(1) Analyzes the characteristics of data model which is suitable for map generalization. Study on the effect of semantic information in map generalization. Introduces a semantic network, and analyzes the topological structure characteristics and the content of knowledge representation of semantic network.(2) Achieve the semantic network representation of map information. Study on the semantic connotation of map basic information. Aiming at the problem that the representation of map information content is complex, using the idea of hierarchical framework, divide the classification of map basic information and build its hierarchy. In view of the situation that the each level of information content of map basic information are different, designing suitable semantic network structure to solve the formal description of map basic information. Through the construction of various types of semantic network, achieve completely information organization of map based on semantic network.(3) Achieve the semantic network representation of map generalization. Through the research on the semantic connotation of map generalization rules and classification analysis, aiming at the problem that map generalization rules are mostly described by natural language, establishing the syntactic pattern of generalization rules. On this basis, build a variety of semantic sub-networks of map generalization rules. According to the problem of multi-scale representation of map generalization, map generalization results are expressed in the form of target scale map. It is implemented by the combined effect of the semantic sub-network for map basic information and the semantic sub-network for generalization rules. Ultimately achieve the realization of the map generalization data model based on semantic network. Map information and generalization rules are integrated to express in this model. And the data storage strategy is given according to the characteristics of the model.The purpose of this paper is to study how to make full use of semantic information of map generalization to improve the degree of automation of map generalization and find a new possible way to completely describe the content of map generalization knowledge. It has a certain significance to enrich the theory and methods of map generalization.