Gualtieri, AndreaSaccà, DomenicoTalia, Domenico2012-11-092012-11-092012-11-09http://hdl.handle.net/10955/62Tesi di dottorato di ricerca in Ingegneria dei sistemi e informatica, XIX cicloThe focus of this thesis is fixed on semantic business process management, aiming to offer an ontology-based support on planning, development and ex post analysis of Information System (IS) based on heterogeneous components. Starting from bases existing in literature, novel methodological and ontological instruments has been defined to represent organizational knowledge ad to support the whole BPM lifecycle, from preliminary analysis of scenario to an ex post evaluation based on the logs generated by heterogeneous functional component. Our contributes address problems that fall into the follow contexts: Strategic BPM: a process-driven methodology for continuous information system modeling has been formulated to support the whole information system life-cycle, from planning to implementation, and from usage to re-engineering. The methodology is based on UML diagram and it is organized on three model layers, to 8 offer a shared view for business analyst and IT engineer both in the scenario evaluation and in the functional setting up and development. Ontological foundation: a framework for an abstract representation of organizational knowledge has been defined to obtain a semantic model to which heterogeneous semantic context can be linked. This framework constitutes an enterprise ontology in which a functional view of organizations is annotated with respect to a conceptual view of abstract topics, allowing an approach that results more flexible than traditional ones to model a dynamic context of a real enterprise. Adopting this framework, independent operations performed by semantically heterogeneous component can be framed as steps of a unique process. Business Process Modelling: a novel approach to model processes and their workflow is obtained by extending flow-oriented standard metamodel with respect to constructs oriented to define process taxonomies. We implement this metamodel adopting a formalism based on Disjunctive Logic Programming extended by objectoriented features to enable knowledge inference on dynamic structures of the process. Business process configuration and execution: a general purpose solution based on a Service Oriented Architecture approach has been planned and applied to semantically integrate heterogeneous tools. By adopting an Enterprise Service Bus we obtained a physical structure able to offer: (1) on the content level, a ontology-driven conceptualization of local data, useful to obtain a synchronization on tools operating to different context; (2) on the dynamic level, a re-organization of local operation logs with respect to category of process belonging to a enterprise process ontology. A same approach allows to supply an enterprise knowledge base useful for analysis purpose. Business process analysis: starting from the enterprise knowledge base generated by logs of execution of distributed and independent tool, new level of knowledge are extracted by adopting process mining and reasoning techniques. In the first case, process schemas have been extracted, analyzing logs with respect to characteristics of the procedures, sorts of users and temporal horizon. In the second case, hidden procedural knowledge has been discovered and taxonomic structures for the execution pattern classification have been obtained, applying logic reasoning based on Disjunctive Logic Programming, to the process schemas expressed in terms of logic rules.enOntologia - ModelliThesis