基于距离的地理信息语义相似性模型优化方法研究

Study on the Optimization of Geo-information Semantic Similarity Model Based on Distance

相似性通常利用基于专家经验的主观判断方法和基于计算机语言学的客观定量描述方法，对概念之间的关系进行定义。在地理信息本体融合共享过程中，通过分析判断地理信息多语义性之间的关系，借助数学理论和方法，建立语义相似度量模型、实现语义相似程度的量化，是解决本体语义异质问题的根本技术途径。现有的语义相似性模型和算法在计算机网络信息查询检索、大数据挖掘、云计算处理、人工智能等领域已有广泛的实际应用，但是针对地理信息的语义相似性算法模型还存在主观性太强、片面性和理想化等不足。本文针对基于距离的算法模型，提出了改进优化的算法思路，提取相似性计算参数，并进行验证，为解决相同领域不同地理信息本体之间的语义信息转换提供算法支撑。

Similarity usually uses subjective judgement method based on expert experience and objective quantitative de- scription method based on computer linguistics to define the relationship between concepts. In the process of geographic informa- tion ontology sharing and fusion, based on the analysis and judgement of the relationship between geographic information seman- tic, establishing semantic similarity measurement model and realizing the quantitative of semantic similarity degree with the aid of mathematical theory and method is the fundamental way to solve the problem of ontology semantic heterogeneity. Existing semantic similarity models and algorithms are widely applied in computer network information retrieval, data mining, cloud computing, ar- tificial intelligence and other fields, but the geographic information semantic similarity algorithm model has some deficiencies, such as too subjective, biased, idealized and so on. In this paper, an improved and optimized algorithm model based on distance is proposed and similarity calculation parameters are extracted and validated, which will provide algorithm support for semantic in- formation transformation among different geographic information ontology in the same field.