摘要
通过对传统语义相似性度量方法的分析和比较,利用相似性关系本身所具有的相对性和传递性,选取共同的领域本体作为相似性比较的基准,提出了基于相似性特征的Web服务语义匹配算法,解决了现有Web服务语义相似性度量方法中需要外在文本集、无法对相似性进行定量表达以及相似性区分程度不明显等问题。最后,选取德国人工智能研究中心发布的基于OWL-S的geography领域服务测试集进行语义相似性度量对比实验,实验结果证明了所提算法的可行性和有效性。
The algorithm for semantic matching of service based on characteristics of similarity,such as relativity and transitivity,was proposed through analyzing and comparing the traditional measurements of semantic similarity and selecting the same domain ontology as the comparison benchmark of similarity.And the proposed algorithm resolved the existing problems in the traditional measurements,such as it need corpus,the semantic similarity measure can’t be quantitative expression,the differences between similarities weren’t obvious and so on.Finally,the services were selected among the geography domain service test set released by the German Research Center for Artificial Intelligence.The results of semantic matchmaking experiment for service proved the validity and the operability of the proposed algorithm.
出处
《测绘科学技术学报》
北大核心
2012年第2期149-152,共4页
Journal of Geomatics Science and Technology
基金
江苏省科技计划项目(BM2010509)
关键词
WEB服务
相似基准
相对性
传递性
服务匹配
Web service
benchmark of similarity
relativity
transitivity
match of service
