The Study Of Natural Language Fuzzy Spatial Relationship Recognition And Simulated Expression

Natural language containing abundant geographical spatial information is closely related to people's daily life.As an indispensable part of spatial information,spatial relations are important for people's cognition of geographical space.As qualitative description,natural language has the characteristics of fuzzy,uncertainty,flexibility and so on.Meanwhile,spatial relations possesses the characteristics of complexity and diversity.With the combination of both,the difficulty of comprehensive is greatly increased.Since the formal language of computer requires a specific standard,how to make the computer identify and understand natural language intelligently as human being is one of the important researches in geographic information science.Taking the natural language as a breakthrough point,this paper explores natural language spatial relationship recognition,description and visualization under fuzzy semantics and realizes the goal of "text-to-figure" transformation.Firstly,through collecting and organizing a lot of references both domestic and abroad,this paper analyzes the existing problems of this field such as the slow progress and lack of depth in the academic research and elaborates the concepts and processes of"text-to-figure".According to different level of "vagueness",description can be divided into three categories,namely completely fuzzy description,interval fuzzy description and quantitative fuzzy description,so as to study the natural language description based on different categories and levels.Secondly,by testing people with different environment familiarity through cognitive experiment,this paper concludes that people get to cognize spatial relations from low level to high level following the sequence of distance relations,direction relations and topology relations,this paper also analyzes how people use natural language to describe spatial relations.Through using descriptive statistics analysis method to classify fuzzy semantic vocabularies,a series of common syntactic patterns concerning topology,direction,distance and direction+distance relationships are built up for further research of semantic parsing and formalized expression.Studying the regularities that how people transfer spatial cognition and descriptions into graphs,this paper defines a couple of indicators through geometric features and uses decision tree algorithm to train the experiment results in order to obtain qualitative distance formal expression rules and build eight direction relation model.Furthermore,this paper puts forward two methods of visualization expression——random parameter method and the method of buffer,which realize the visualization of natural language spatial relations descriptions.Finally,in the example of Xianlin campus of Nanjing Normal University,the system adopts a three-tiered architecture containing user layer,service layer and data layer,designs and develops the analogous expression system of natural language spatial relations,which realizes the "text-to-figure" conversion between fuzzy natural language description and point-line features,and verifies the validity of the method proposed in this paper.Most importantly,this system contributes a lot to recognition,reasoning,calculation,formal expression and visualization expression of geographical spatial information.