Contextualized and Personalized loCation-based serviCes

摘要: Advances in the technologies of smart mobile devices and tiny sensors together with increase number web resources open up a plethora new information services where people can acquire disseminate at any place time. Location-based (LBS) are characterized by providing users useful local information, i.e. that belongs to particular domain interest user be use while remains area. In addition, LBS need take into account interactions dependencies between services, context for filtering delivery order fulfill needs constraints users. We argue consequently it brings series technical challenges terms data semantics infrastructure, context-awareness personalization, as well query formulation answering etc. They not simply extended from existing traditional management strategies. Instead, they solution. Firstly, we propose semantic infrastructure on basis modularized ontologies approach. elaborate core ontology which is mainly composed three modules describing contexts. The aims presenting an abstract view (a model) all LBS. contrast, instances (of actual contextual data) stored independent stores, called service profiles, profiles profiles. These semantically aligned concepts through set mappings. This approach enables distributed sources maintained autonomous manner, adapted high dynamics mobility sources. Secondly, separately address function, features, our modelling major players, service, Then, define constructs represent their inter-dependencies illustrate how these contribute personalized contextualized processing. Service classes organized taxonomy, distinguishes business functions. concept hierarchy helps analyze reformulate users' queries. introduce kinds relationships module enhance services. identify five key components contexts regard them Component related specific composition describe spatio-temporal constraints. A profile contains personal about given possibly self-defined rules, offer hints what likes or dislikes, could attract him her. clustering common features help cooperative answering. Each itself. inter-related link belonging two different modules. fully benefits structure ontology. It allows restricting search space, facilitating maintenance each module. Finally, studied reformulation processing issues How make interface tangible provide rapid relevant answers typical concerns Our format only obeys "simple, effective" golden-rules user-interface design, but also satisfies domain-independent emphasizes importance With pre-defined operators, easily specify queries availability looking for. eliminating most irrelevant usually generated keyword-based approaches. Constraints various dimensions (what, when, what-else) expressed conjunctive query, then smoothly translated RDF-patterns. strategy using SPARQL syntax, explain relaxation done rules specified profile.