基于地名本体的加权语义相关度算法研究  

Research on the Weighted Semantic Correlation Algorithm Based on Toponym Ontology

作　　者：索俊锋[1] 郑海晨[1] SUO Jun-feng;ZHENG Hai-chen(School of Civil Engineering,Northwest Minzu University,Lanzhou,730124,China)

机构地区：[1]西北民族大学土木工程学院

出　　处：《西北民族大学学报（自然科学版）》2019年第3期1-8,25,共9页Journal of Northwest Minzu University(Natural Science)

基　　金：甘肃省自然科学基金项目(18JR3RA367)

摘　　要：针对现有地理数据空间检索方法以简单字面匹配为核心、缺乏推理能力等问题,根据地理信息科学中的空间拓扑理论,提出一种基于地名本体的加权相关度算法.该算法考虑到不同的空间拓扑关系在语义检索应用中的重要程度不同,从而赋予地理空间本体关系不同的权值,并体现了其与语义距离的反比关系.通过地名本体实例验证,表明经加权后的语义相关度算法在语义检索应用中可以达到更好的检索效果.On the basis of on the some problems in the present geographical spatial data retrieval method,such as with a simple literal matching as a core,lacking of reasoning ability,a weighted correlation algorithm based on toponym ontology was put forward according to the space topological theory of geographical information science.The algorithm considered that the different spatial topological relations in the semantic retrieval application had the different importance,and was assigned the different weights in accordance with the spatial relationship importance of geographic ontology,meanwhile.It also reflected the inverse relationship with semantic distance.The present study had proved that the weighted correlation algorithm could achieve a better retrieval effect in the application of semantic retrieval through the example of the toponym ontology.

关 键 词：地名本体 相关度 语义检索 语义距离 权重 拓扑关系 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]