A Ranking Mechanism for SemanticWeb Service Discovery

This paper presents an approach for ranking semantic Web service advertisements with respect to a service request. The use of recall and precision is proposed as suitable measures for determining the degree of match between the request and the advertisement. Ranking is based on the use of the domain ontology to infer the semantic similarity between the parameters of the request and the advertisement. The proposed approach is applicable to several types of ontologies, ranging from simple taxonomies to highly expressive ontologies, such as OWL ontologies.