Process Event Streams: What You Need To Know (Part 1)

Jan 7, 2011   //   by John   //   Business Process Management  //  2 Comments

A dictionary defines an event as “something that takes place; an occurrence”. Therefore, a business process event represents anything that happens or is happening in a business process. When processed or captured by a computer system, an event object, entity or data structure represents or records such an event. An event stream is a sequence of events, usually ordered by time.

Before discussing business process events in particular, their relationship to simulation, and the standards that are available to represent events and event streams, it is helpful to understand the kind of applications that rely on event streams. There is a lot of terminology in this application space, and a lot of the terms happen to overlap. Here is how I understand the main application areas:

Process Mining

This is the act of using system event logs to analyze business processes. Event logs generated by computer systems can often be used to extract knowledge about the processes these systems support:

Process mining aims at improving this effort by providing techniques and tools for discovering process, control, data, organizational, and social structures from event logs.

Source: Eindhoven University of Technology Business Process Management Center

Process mining can be used for process discovery. In this case, there is no existing description of the business process, so event logs are used to ascertain the structure or a description of the business process. When a description of the process already exists (for example a process model or diagram), process mining can be used to test conformance between the actual process and the pre-existing description. Event logs are used to find and report discrepancies between real world processing and the process description. Lastly, process mining may be used to extend a process description, by finding discrepancies and then using them to modify or update the process description to reflect the process as it actually occurs.

Business Process Intelligence (BPI)

BPI, or simply process intelligence, is also known as operational intelligence. It is the real time monitoring of business processes and/or specific activities within a business process. This kind of monitoring allows management to react to service interruptions, bottlenecks and other inefficiencies, as well as potential risks and threats as they occur. BPI is achieved by capturing process event data and calculating metrics based on these events in real time. Typically the resulting data is published on dashboards, or used to send notifications of exceptional circumstances via some messaging channel (e.g. email, SMS messages, etc.).

Complex Event Processing (CEP)

CEP is the underlying technology behind many process intelligence solutions and involves the monitoring, recording and filtering of events to identify meaningful “high level” events within the event cloud (a collection of event streams). By monitoring and correlating many events happening across all the layers of an organization, complex events can be inferred. For example, late deliveries plus an abnormal rate of absenteeism might infer bad weather or a traffic disruption near an operation. By analyzing the impact of complex events, action can be taken in real time (e.g. reroute deliveries, contact employees to suggest leaving early for work).

Business Activity Monitoring (BAM)

This is a software solution that provides a real-time summary of business activities to management. This is done by aggregating, analyzing and presenting real time information on activities in the business process, usually on some kind of dashboard that provides a set of key performance indicators (KPI’s). Management can use this to measure performance,¬†adherence,¬†as well as to ascertain potential problems. All BAM systems process events. Some BAM systems are based on CEP which allows them to present information on higher level complex events which, as described in the previous section, are identified by correlating several lower lever events in the event stream/cloud. Clearly, these CEP based systems blur any distinction between BAM and BPI (frankly, BPI without CEP seems awfully similar to BAM as well for that matter).

All of these applications rely on streams of business process events. Event logs are just event streams persisted in some format, and event clouds are simply a collection of event streams – typically generated by one or more distributed systems. Strictly speaking, event clouds are not necessarily an ordered set of events (a stream implies some kind of order, usually temporal).

Coming soon: More on events, the relationship between business process event streams and simulation, and standards used to define events and event streams.

More reading:


Leave a comment