Complex Event Processing (CEP) is the discipline of detecting, filtering, correlating, aggregating and processing events in real-time, which have potentially complex relationships with other events and artifacts of interest.

Key Concepts

• Events are occurrences of interest
• Events are technology neutral
• Events have context
• Events can carry metadata about themselves
• Events can be evaluated based on event data, event metadata (carried with the event) or event context (other event metadata)
• Events form spatial, temporal or other types of patterns
• Events represent ad hoc processes, i.e. a sequence of occurrences that represent a process and/or result in the execution of a process (e.g. a revenue recognition process might consist of “letter of intent signed”, “contract signed”, “purchase order received”, “product shipped” and “product delivered and accepted” events)
• Events may be nested (e.g. a “change of address” event followed by an “ATM card lost” event could represent a compound event called “attempted fraud”)
• Events can optionally generate responses or actions

Properties of Complex Event Processing

Unlike traditional process automation solutions, CEP is characterized by:

• Nonlinear, often unpredictable activity flows
• Bottom-up process control
• Dynamic logic
• Execution via event management rather than rules and workflow
• Processes defined by “borders” using exception management
• Closed-loop processes, with outcomes consumed by the next iteration for continuous adjustment
• Time, sequencing, and affinity as integral components of evaluation

Event Stream Processing (ESP)

Event Stream Processing (ESP) engines process event streams to identify events of interest and derive information from these events in real-time. ESP may therefore be regarded as a subset of CEP to the extent that complex events are involved in these event streams. ESP may also be regarded as distinct from CEP to the extent that simple events are involved. ESP and CEP both involve real-time event processing that results in an Event-Driven Architecture (EDA). History

Professor David Luckham of Stanford University is known for developing CEP concepts.

Governance and Risk Management Implications

Complex Event Processing (CEP) concepts are sometimes applied by governance, risk management, and compliance (GRC) solutions for detection and response of complex event sequences. As real-time processing requirements increasingly impacts GRC solutions spanning organizational and risk management silos, CEP capabilities will become more important.

References

Complex Event Processing