Web Services Discovery in A Pay-As-You-Go Fashion
Ying Pan,Yong Tang,Shu Li
DOI: https://doi.org/10.3217/jucs-017-14-2029
2011-01-01
Abstract:Extensive effort has been brought forth to assist in web service discovery. In particular, classic Information Retrieval techniques are exploited to assess the similarity between two web services descriptions, while Semantic Web technologies are proposed to enhance semantic service descriptions. These approaches have greatly improved the quality and accuracy of service discovery. However, these works require hard up-front investment before offering powerful functionalities for service discovery, and they do not study how to discover web services in a pay-as-you-go fashion. In this paper, a framework based on dataspace techniques is proposed to discover web services in a pay-as-you-go fashion. In this framework, a loosely structured data model based on dataspace models is presented to describe web services and the relationships among them, and then keyword-based query is supported on top of this model by using the existing dataspace query language. To support similarity-based service discovery, dataspace techniques are extended to declare the similarity among web services, and a discovery algorithm is presented. In addition, a lightweight way adding semantics to the query processing is also shown in the paper. Finally, the differences between our work and previous works are discussed.