Abstract: Because of semantic absence in the traditional web, the services composition efficiency and accuracy are greatly reduced. The Semantic Web overcomes the lack of semantic in the traditional web. Since there exists abundant knowledge in the Semantic Web, it becomes more and more important to effectively utilize these semantics to achieve services composition. In order to solve this problem, this paper presents a dynamic service composition mechanism based on the multidimensional user model. It realizes the sequential and parallel service composition through the semantic matching. Adopting this mechanism will improve the Semantic Web Services composition efficiency and accuracy to some extent.

Key words: Service Composition, Sequential Composition, Parallel Composition, OWL-S, Description Logic

摘要： 传统Web服务的复合由于缺乏足够的语义信息而大大降低了服务复合的效率和准确率。语义Web弥补了传统Web语义不足的缺点。但是，面对语义Web上分布的海量知识，如何有效地利用语义信息实现Web服务复合成为一个迫切需要解决的问题。文章针对这一问题，提出一种基于多维用户模型的动态服务复合机制。该机制在Web服务语义标识的启发下通过语义匹配算法能够实现Web服务的顺序复合和并行复合。采用这种服务复合机制将会在一定程度上提高语义Web服务复合的效率和准确率。

关键词: 服务复合, 串行复合, 并行复合, OWL-S, 描述逻辑