A provenance-based approach to semantic web service description and discovery

摘要

Web services have become common, if not essential, in the areas of business-to-business integration, distributed computing, and enterprise application integration. Yet the XML-based standards for web service descriptions encode only a syntactic representation of the service input and output. The actual meaning of these terms, their formal definitions, and their relationships to other concepts are not represented. This poses challenges for leveraging web services in the development of software capabilities. As the number of services grows and the specificity of users' needs increases, the ability to find an appropriate service for a specific application is strained. In order to overcome this challenge, semantic web services were proposed. For the discovery of web services, semantic web services use ontologies to find matches between user requirements and service capabilities. The computational reasoning afforded by ontologies enables users to find categorizations that weren't explicitly defined. However, there are a number of methodological variants on semantic web service discovery. Based on e-Science, an analog to e-Business, one methodology advocates deep and detailed semantic description of a web service's inputs and outputs. Yet, this methodology predates recent advances in semantic web and provenance research, and it is unclear the extent to which it applies outside of e-Science. We explore this question through a within-subjects experiment and we extend this methodology with current research in provenance, semantic web, and web service standards, developing and empirically evaluating an integrated approach to web service description and discovery. Implications for more advanced web service discovery algorithms and user interfaces are also presented.