基于统计式机器学习的地理本体融合模型

Geography Ontology Fusion Model Based on Statistical Machine Learning

摘要:

针对不同领域对地理事物的认知体系差异造成了地理本体异构的问题,提出了地理本体融合模型,引入统计式机器学习的方法对概念间的关系进行自动处理,并以概念间关系在不同本体出现的频度来产生其可信度,最后形成带有统计信息和领域信息的大型地理概念空间.该模型巧妙规避概念层面繁琐的异构映射过程,融合概念空间将多个地理本体所表达的概念知识融为一体,并保持了领域内的信息,有效实现了不同认知体系之间的共享.

Abstract:

The ubiquity of geography ontology heterogeneity is caused by multiple cognitive systems for geographical object.Geography ontology fusion model was proposed by introducing the method of automatic statistical machine learning for processing relationship within the concepts.The credibility was produced according to emergence frequency of relationship between concepts in different ontology and finally a large-scale integrated geographic concept space with statistic and field information was generated.Cumbersome concept mapping process is circumvented through this model,and all the knowledge expressed in ontologies is fused while information within each field is preserved in this concept space which realize sharing between multiple cognitive systems.