Session-based dynamic interaction models for stateful web services

The prevalence of the service paradigm spans diverse domains like commercial, social, technological or scientific. Due to its simplicity and familiar semantics, it provides a powerful general abstraction for system programming, interaction, and integration. Several standardisation efforts have further contributed to the popularity of the service concept and its usage, since this provides a uniform access to and aggregation of entities with different characteristics and at different levels of the cyberinfrastructure. The perceived current trend on making everything accessible as a service (XaaS) builds on such service characteristics, and examples range from Web-enabled Wireless Sensor Networks, the Internet of Things and Web of Things, to Cloud computing, and the Internet of Services. Upon the acknowledgement of such high heterogeneity and of the extreme large scale of emerging service systems, Service Science presents a novel and overarching view on analysing and developing further the service paradigm. The high complexity of current and future service systems in this domain, require innovative solutions to be developed in order to improve service productivity and quality. To this extent, this work concentrates on service engineering Web services and proposes a solution based on the Session concept contributing to solve open problems on service system interaction and adaptation. The focus is on the interactions between Web services interfacing stateful resources and its clients, in particular. The session abstraction is used to: a) capture the service/users interaction context, b) support dynamic interaction models within, and c) contextualize on demand and automatic dynamic adaptations. The major goal is to capture Web service/users interactions modelled as Sessions, in order to simplify their re-use and adaptation in the context of the cited Services Sciences' complex systems.