Refine
Document Type
- Article (2)
- Master's Thesis (1)
Has Fulltext
- yes (3)
Is part of the Bibliography
- no (3)
Keywords
- CEP (3) (remove)
Institute
- Fakultät IV - Wirtschaft und Informatik (3) (remove)
Heterogeneity has to be taken into account when integrating a set of existing information sources into a distributed information system that are nowadays often based on Service- Oriented Architectures (SOA). This is also particularly applicable to distributed services such as event monitoring, which are useful in the context of Event Driven Architectures (EDA) and Complex Event Processing (CEP). Web services deal with this heterogeneity at a technical level, also providing little support for event processing. Our central thesis is that such a fully generic solution cannot provide complete support for event monitoring; instead, source specific semantics such as certain event types or support for certain event monitoring techniques have to be taken into account. Our core result is the design of a configurable event monitoring (Web) service that allows us to trade genericity for the exploitation of source specific characteristics. It thus delivers results for the areas of SOA, Web services, CEP and EDA.
This article discusses event monitoring options for heterogeneous event sources as they are given in nowadays heterogeneous distributed information systems. It follows the central assumption, that a fully generic event monitoring solution cannot provide complete support for event monitoring; instead, event source specific semantics such as certain event types or support for certain event monitoring techniques have to be taken into account. Following from this, the core result of the work presented here is the extension of a configurable event monitoring (Web) service for a variety of event sources. A service approach allows us to trade genericity for the exploitation of source specific characteristics. It thus delivers results for the areas of SOA, Web services, CEP and EDA.
Das Bedürfnis Daten in Echtzeit zu analysieren und auf Ereignisse zu reagieren, ist innerhalb aller Branchen in den letzten Jahren stark gestiegen. Als die Analysetechnik für Echtzeitdatenströme hat sich das Complex Event Processing (CEP) durchgesetzt. Mithilfe von Regeln lassen sich kausale, temporale und räumliche Zusammenhänge von Ereignissen definieren und durch eine CEP-Engine evaluieren. Die Konstruktion von Regeln hat sich dabei als einschränkende Faktor von CEP herausgestellt. Greedy4Cep ist ein algorithmischer Ansatz zur automatisierten Erstellung von CEP-Regeln anhand eines historischen Datenstromes.