摘要
将功能相似的Web服务聚类是一种有效的服务发现方法,而聚类的关键在于服务之间语义相似度的计算。目前国内外主流使用关键字、信息检索和基于本体的方法计算相似度,这些方法存在语义信息缺失等问题,并且聚类方法只考虑到简单数据类型的处理。本文提出一种同时包含处理简单数据类型和复杂数据类型的本体学习方法,利用本体学习和信息检索相结合的方式(Hybrid term similarity,HTS)进行Web服务聚类。实验结果表明,该方法能够有效地提高Web服务的聚类效果。
Establishing Web services into function similarity cluster is an efficient method of service discovery. The key of the clustering is the calculation of the semantic similarity between Web services. Mainstream use keywords, information retrieval or ontology-based method to compute the similarity in home and abroad. Furthermore, These methods exist such problems as lack of semantic information. Further, current clustering methods only take into account the processing of simple data type.The approach is proposed to calculate the service similarity not only contains simple data types but contains complex data types. Thus, use ontology learning and information retrieval method to Web service clustering. This approach used in project will significantly improve Web services discovery.
出处
《电子设计工程》
2016年第19期11-14,共4页
Electronic Design Engineering
基金
陕西省教育厅科研项目(12JK0733)
陕西省自然基金项目(2012JM8044)
西安邮电大学研究生创新基金项目(114-602080049)
