摘要
自然语言使用自由、灵活,并且不受环境和地点的约束,是一种流畅、快捷的信息交流方式。在地理信息系统中合理地运用自然语言,实现用户与系统之间的人机交互,将有助于G IS的智能化发展。自然语言以文字的形式发送G IS命令,其中负载了大量空间数据信息和G IS的基本操作指令。因此,基于受限自然语言的G IS命令解析,旨在从输入的语句中抽取与G IS操作有关的信息。本文根据规则匹配的思想,按地图浏览、数据读写、数据采集、信息查询、空间分析功能和词语搭配关系对词汇进行分类标记,并且通过语句切分、语句规范化、规则匹配及函数映射四个处理过程,对以受限自然语言为输入形式的G IS命令进行解析。经实例分析表明,规则匹配方法可以较好地消除自然语言中非结构化空间信息与G IS中结构化空间信息之间的语义障碍,有助于自然语言在G IS中的深入应用。
Natural language can be applied freely and flexibly, which would not be restrained by the environment and locations, and it is a fluent and quick communication method. Thus, reasonably applying natural language in （GIS） to realize the human-computer interaction between users and system will contribute to the development of GIS intellectualization. Natural language will transmit the GIS commands in the form of the characters, in which a vast number of spatial data and information as well as GIS basic operation instructions are loaded. Therefore, GIS command analysis based on constrained natural language aims to extract the information related to GIS operations from input sentences. According to the thought of rule matching, the essay marks the words with classifications, based on map browsing, data reading/writing, data collections, information inquiry, spatial analysis functions and word collocation relations. With sentence segmentation, sentence normalization, rtde matching and function mapping to analyze the GIS commands from constrained natural language input. Through case study, it indicates that, the rule matching method can efficiently eliminate the semantic barriers existing between unstructured spatial information in natural language result, it will contribute to further applications of natural language and structured spatial information in GIS. As a in GIS.
出处
《地球信息科学》
CSCD
北大核心
2009年第2期183-188,共6页
Geo-information Science
基金
国家高新技术研究发展计划(863计划)资助项目(2007AA12Z218
2007AA12Z221)
江苏省大学生创新计划项目(1812200013)
关键词
自然语言
GIS
语句
规则匹配
natural language
GIS
sentences
rule matching
