An Extensible and Personalized Approach to QoS-enabled Semantic Web Service Discovery

We present a framework for the autonomous discovery and selection of Semantic Web services based on their QoS properties. The novelty of our approach is the wide use of semantic technologies for a customizable discovery, which enables both the service users and providers to flexibly specify their matching models for QoS and the corresponding environmental conditions. In the presented approach, the discovery and ranking of services can be personalized via the use of domain ontologies detailing the user's preferences and the provider's specification. The discovery component is modeled as an adaptive query processing system in which the basic steps of filtering, matchmaking, reputation-based QoS assessment, and ranking of services correspond to logical algebraic operators, which facilitates the introduction of different discovery algorithms and the automatic generation of appropriate parallelized matchmaking evaluations, enabling the scalability of our solution up to unpredictable arrival rate of user queries against high numbers of published service descriptions in the system.