Evaluation of Semi-Automatic Acquisition of Semantic Descriptions of Web Services

This paper introduces and evaluates a novel Web service interface annotation and matching scheme designed for processing large sets of Web service interface descriptions. The matching scheme relies on a set of rules, which exploit the structural relationships encoded in an automatically learned ontology to discover matching interface elements. The proposed scheme is evaluated on a Web services interface corpus consisting of WSDL descriptions. The experimental results show that the proposed matching scheme can be used for bootstrapping large-scale Web service interface annotation.