Towards unified and native enrichment in event processing systems

Abstract

Events are encapsulated pieces of information that flow from one event agent to another. In order to process an event, additional information that is external to the event is often needed. This is achieved using a process called event enrichment. Current approaches to event enrichment are external to event processing engines and are handled by specialized agents. Within large-scale environments with high heterogeneity among events, the enrichment process may become difficult to maintain. This paper examines event enrichment in terms of information completeness and presents a unified model for event enrichment that takes place natively within the event processing engine. The paper describes the requirements of event enrichment and highlights its challenges such as finding enrichment sources, retrieval of information items, finding complementary information and its fusion with events. It then details an instantiation of the model using Semantic Web and Linked Data technologies. Enrichment is realised by dynamically guiding a spreading activation algorithm in a Linked Data graph. Multiple spreading activation strategies have been evaluated on a set of Wikipedia events and experimentation shows the viability of the approach.

On-the-fly generation of multidimensional data cubes for web of things

“…Therefore, we are researching on techniques that support other types of databases and not just triple store. Event enrichment is an important activity in our approach, therefore, we are also investigating native approaches to event enrichment [9] and approximate semantic event-processing [8] techniques and determining how this would effect on-the-fly cube generation. We are also working on improving the ontology, implementation and our methodology to generate data cubes on-the-fly in minimum time possible.…”

When things matter: A survey on data-centric internet of things

“…Teymourian et al [2012b] discuss the benefits of background knowledge for event processing and describe different categories of event query rules. Hasan et al [2013] identify an information completeness problem in semantic event processing contexts from a different angle. For example, while the basic information item in an event-based system is an event, normal users often require the system to handle information that is not encoded in the event.…”

“…Several challenges are identified for event enrichment, including determination of the enrichment source, retrieval of information items from the enrichment source, finding complementary information for an event in the enrichment source and fusion of complementary information with the event. To address these challenges, a model based on unifying enrichment within the event consumer logic and a native enricher that tackles incompleteness before matching are proposed [Hasan et al 2013].…”

Approximate Semantic Matching over Linked Data Streams

“…The experiments have been conducted using real-world data, which is a set of events extracted from DBpedia, provided by the authors of the work in [20]. We used these events in RDF format to form a Linked Data stream so as to simulate the sensing data streaming process in the smart city scenario.…”