From situation modelling to a distributed rule-based platform for situation awareness: an ontological framework for disaster management applications

摘要

Situation-aware (SA) applications are particularly useful for disaster management. The complex nature of emergency scenarios presents challenges to the development of collaborative and distributed SA solutions. These challenges concern the whole lifecycle, from specification to implementation phases, such as how to model the reaction behavior of a detected situation and how to provide an interoperable situation notification service. In addition, treating unforeseen situations, i.e. situations not modeled at design-time, is an open topic. In this PhD research we intend to deal with these issues by proposing a framework for SA application development, which is applicable in different emergency scenarios and covers the conceptual, specification and implementation aspects, through an ontological model-driven methodology. The conceptual requirements can be addressed by adopting an appropriate foundational ontology. The specification aspects include the design of the context, the situation types of interest and the decision to be performed by the application when a situation is detected. For supporting the implementation of the SA application, we intend to improve the interoperability and performance of an existing rule-based situation notification platform, which is based on service oriented architecture and complex event processing. A separate module takes advantage of predictive analysis techniques to detect unforeseen situations. We plan to evaluate our framework with SA application prototypes and proofs-of-concept.