An automated workflow composition to semantic web services

Nowadays, application and software realized under Service-Oriented Architecture using Web Service technology is popular. To fulfill user´s requirements, usually, several web services must be composited as a more complex, value-added service. As more and more web services are available, it is very hard to compose services manually. In our previous work, an approach, which is able to simultaneously compose workflow and consider global quality degree of solution service, has provided and its cornerstone is case-base reasoning as well as genetic algorithm. In this paper, we extend our previous work to address this issue via concept from semantic web, namely semantic web service, that makes web service can be operated automatically by machine. The inability of calculation to truly similarity of two elements must be eliminated in order to ensure that our approach covers all of the dimensions of service composition problem. To this goal, ontology and semantic symbols are introduced to rewrite the original definitions of fitness function of genetic algorithm, making workflow with stateful and semantic-based dataflow.