语义相似性算法因子评价方法研究  

Research on the Evaluation Method Based on Algorithm Factors of Semantic Similarity

作　　者：熊顺[1,2,3] 刘平芝[2,3] 张卫柱[2,3] 金澄[1,2,3] 

机构地区：[1]信息工程大学地理空间信息学院,河南郑州450052 [2]西安测绘研究所,陕西西安710054 [3]地理信息工程国家重点实验室,陕西西安710054

出　　处：《测绘科学与工程》2014年第3期68-72,共5页Geomatics Science and Engineering

摘　　要：地理信息要素语义信息的智能化转换，很大程度上依赖于语义相似性算法因子的正确判断。本文利用模糊数学综合方法评价语义相似性算法因子的影响大小，充分考虑到了评价等级之间的模糊性、过渡性；对影响语义相似度的因素进行科学分解，先对其进行平行评价，然后对各因素的影响结果再进行综合评价。既反映了各影响因素之间的差异性，又体现了它们之间的相关性，具有方法科学、步骤严谨、结果可靠、符合实际的特点。The intelligent conversion between the semantic information of geographic information elements mostly depends on the accurate judgment of the semantic similarity algorithm factors. In this paper, the fuzzy mathematics comprehensive method is used to evaluate the influence of semantic similarity algorithm factors. The method takes full account of the ambiguity and transi- tivity between different evaluation levels, and decomposes the factors that influence semantic similarity in a scientific way. Firstly, all the factors are evaluated in parallel, and then the influences of each factor are evaluated comprehensively. In this way, both the differences and correlation between each factor are reflected. It proves that the method is scientific, rigorous, reliable and practical.

关 键 词：模糊数学 语义相似性 评价方法 算法因子 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]