摘要
针对现有多数语义Web服务发现方法应用实施难度大,对终端用户输入信息的完整性依赖度高的问题,提出一种基于简单查询语句的轻量级语义Web服务发现模型。该模型提供给用户一个类似Google的查询界面,输入查询语句后经过领域本体匹配、基于WordNet同义词典匹配等步骤自动发现并调用相应的Web操作。另外,还利用自学机制不断扩充本体词汇以提高系统的准确率和召回率。详细分析了该模型的系统性能,并深入研究了不同情况下准确率和召回率的变化。实验结果表明,本体匹配技术及自学机制的使用是系统准确率和召回率提高的关键。
Most methods of the current semantic-based web service discovery are difficult to be implemented and are too dependent on the information inputted by end user. In order to solve？ the aforementioned？ problem, a model for lightweight semantic web service discovery based on user query is proposed, in which a user query interface like Google search engine is provided and the web service opera- tions will be discovered automatically through the process of domain-specific ontologies matching and thesaurus matching based on Word- Net dictionary after the user query is inputted. What＇ s more, the self-learning mechanism is used to enrich ontologies vocabularies in order to enhance the precision rate and the recall rate. The system performance is analyzed in details and the changes of the precision rate and the recall rate under different conditions is researched. The experiments show the methods of domain-specific ontologies matching and self-learning mechanism are vital to increase the precision rate and the recall rate.
出处
《计算机工程与设计》
CSCD
北大核心
2011年第2期450-452,456,共4页
Computer Engineering and Design
基金
河南省科技厅基础与前沿技术研究基金项目(092300410160)
河南省重大科技攻关基金项目(092102110274)
关键词
语义WEB服务
服务发现
领域本体
WORDNET
自学机制
semantic web service
service discovery
domain-specific ontologies
WordNet
self-learning mechanism
