Semantic Complex Event Processing for Decision Support

An increasing amount of information is being made available as online streams, and streams are expected to grow in importance in a variety of domains in the coming years (e.g., natural disaster response, surveillance, monitoring of criminal activity, and military planning [7,22]). Semantic Web (SW) technologies have the potential to combine heterogeneous data sources, leveraging Linked Data principles, but traditional SW methods assume that data is more or less static, which is not the case for streams. The SW community has attempted to bring streams to a semantic level, i.e., Linked Stream Data, and a number of RDF stream processing engines have been produced [1,4,13,20]. This thesis work aims at developing and evaluating techniques for creating aggregated and layered abstractions of events. These abstractions can be used by decision makers to create better situation awareness, assisting in identifying decision opportunities, structuring and summarizing decision problems, and decreasing cognitive workload.