METEOR-S Web Service Annotation Framework with Machine Learning Classification

Researchers have recognized the need for more expressive descriptions of Web services. Most approaches have suggested using ontologies to either describe the Web services or to annotate syntactical descriptions of Web services. Earlier approaches are typically manual, and capability to support automatic or semi-automatic annotation is needed. The METEOR-S Web Service Annotation Framework (MWSAF) created at the LSDIS Lab at the University of Georgia leverages schema matching techniques for semi-automatic annotation. In this paper, we present an improved version of MWSAF. Our preliminary investigation indicates that, by replacing the schema matching technique currently used for the categorization with a Naïve Bayesian Classifier, we can match web services with ontologies faster and with higher accuracy.