A Research On Acquisition Of Geographical Knowledge Between Geographical Entities Based On Semantic Grammar

Geographic information and data are important components of the objective knowledge world. Various relationships among geographic entities on the Internet are in the form of unstructured text, how to use an intelligent way to acquire knowledge from countless text online is a very urgent issue in current field of knowledge acquisition. And at present,the modern research on spatial relationship goes against the sharing and reuse of the spatial relationship knowledge owing to lacking of layer concepts and can't expressing sort relations fully. In recent years, geo-ontology research has been paid more attention by scholars from the fields of GIS. Through absorbing ontology theory, the paper researched how to extract various relationships among geographic entities from unstructured geographic texts, and established spatial relation reasoning mechanism to realize the geographic information sharing and interoperability finally.This paper carried out the following work, including two parts.1. GeoRSG(Geographical Relationship Semantic Grammar) was constructed, which reflects geographic relationships in Chinese written language. GeoRSG also reflects a classification of relationships among geographic entities, and uses a rule-based method to depict linguistic expressions of relationships in the texts. With GeoRSG, we can obtain the composite positional relationships among multiple geographic entities from texts online accurately.2. Implemented a parser, called the GeoRSG Parser, which was described for extracting the geographical knowledge in the form of the predicate with the help of GeoRSG. Experiments indicate that the method can obtain triples relationships among geographic entities, which proportion is 9.11%, and has achieved a precision rate of 89.13%.