Characterization and search of web services through intensional knowledge

Web service technologies are widely adopted to access services and compose new applications starting from software components available from the shelf. Consequently, more and more service descriptions are becoming available on the network to designers, who often filter them according to keyword-based search, thus obtaining huge amounts of matching results, that, if not properly controlled, lead to an information overload that might cause confusion rather than knowledge. In this paper, we propose to apply data mining techniques to SOAP-based service descriptions in order to infer patterns providing a summarized and integrated representation of service functionalities. These patterns provide succinct (intensional) knowledge that can be directly queried or used to drive exploratory searches. Specifically, we propose W-DREAM (Web services DiscoveRy via intEnsionAl knowledge Mining), an infrastructure to perform intensional service representation and querying to support application designers to select the web services that best suit their needs.