Complex event processing is a way of analyzing information about event(s) and gaining insight or conclusions as the event(s) occur. Complex event processing usually combines data from multiple sources to look for trends or patterns. The goal, of course, is to be able to respond quickly to a specific event. I used to refer to these techniques as setting thresholds or alerts.
For example, something that happened to me: I had never bought anything (except airline tickets and hotel nights) with a specific credit card, but over the weekend I used the credit card to buy tires for my Pathfinder. This was out of the ordinary behavior for me, and the credit card company sent up an alert for me to call them and approve the transaction. I thanked them for watching out for fraud on my credit card, and thought how nice is that!
Data quality and integration during event processing can become very complex when we have to use multiple sources of data. It is extremely important for us “data people” to keep it as simple as possible, and remember that perfection is costly in terms of time to market and cost. So, the balance is where we need to be. I usually gather requirements for alerts and thresholds and benchmark them with a proof of concept to make sure the service level agreements can be met.
It is assumed by the experts and analysts that complex event processing will enable organizations to analyze patterns using current information.