An analytical model for structuring natural language based on knowledge unit—Taking domain rules of urban planning as examples

When processing knowledge rules described by natural language in computer, it is necessary to convert non-structured language to structured one. Especially, for the particular domain rules with spatial information, to eliminate semantic obstacle of spatial information between the two kinds of language is a difficulty for transformation. Taking domain rules in urban planning as examples, a concept of knowledge unit is proposed, a knowledge unit system is constituted and the combination patterns of knowledge units are abstracted. Based on them, this paper constructs an analytical model for natural language. Applying this model, rules described by natural language can be structured into combination pattern based on knowledge units, subsequently, be transformed into structured computer language for geography information system(GIS) inquiry, analysis and manipulation.