# Process Mining

**In short:** Process mining is a family of techniques that analyze event data from information systems. It shows how a process actually runs, not how people believe it runs, and helps improve it.

## Explanation

Process mining is a group of techniques. They analyze event data to understand and improve operational processes.

**Where the data comes from.** Every step of a business process leaves a digital footprint in the systems that handle it. This is called event log data. Process mining software extracts it from those transactional systems.

**What an event log needs.** The minimum input is data linked to three things:

- a case ID, which ties events to one run of the process,
- the activities that happened,
- timestamps, which say when they happened.

**Reality, not belief.** The main value is a view of how processes actually run. People often hold a different picture of how they think the process runs. Event data gives a factual basis for comparison.

**Three classes of techniques.** Process mining techniques fall into three main classes: discovery, conformance checking and enhancement.

**Exchanging event data.** The XES standard (IEEE 1849) defines a language to transport, store and exchange event data. Process mining is one example use. The standard was revised in 2023 as IEEE 1849-2023.

**Community.** The IEEE Task Force on Process Mining promotes the topic. It published the Process Mining Manifesto.

For related ideas, see the articles Process Analysis and BPM Life Cycle.

## Example

A support team handles customer tickets in a ticketing tool. Each ticket is a case with its own ID. Each status change is an activity with a timestamp. The team extracts these records from the tool. Analysts then study how tickets actually moved. They compare that with the flow the team believes it follows.

## Trade-offs

Process mining only works when the systems record events with a case ID, activities and timestamps. Without these three, the data cannot serve as input.

BPMN diagram: [Process mining flow: from event logs in systems to a view of how the process actually runs.](https://arc.fuet.ch/diagrams/concepts/modeling-bpmn/process-mining/diagram-1.bpmn)

## Checks

- The document states that event data for process analysis links to a case ID, activities and timestamps.
- If event data is exchanged between tools, the document names the format used, such as XES (IEEE 1849).

## Related

- [Process Analysis](https://arc.fuet.ch/concepts/modeling-bpmn/process-analysis.md)
- [BPM Life Cycle](https://arc.fuet.ch/concepts/modeling-bpmn/bpm-life-cycle.md)
- [AS-IS and TO-BE View](https://arc.fuet.ch/concepts/documentation/as-is-and-to-be-view.md)
- [Process Model versus Process Instance](https://arc.fuet.ch/concepts/modeling-bpmn/process-model-versus-process-instance.md)
- [Runtime View](https://arc.fuet.ch/concepts/documentation/runtime-view.md)
- [BPMN 2.0 Notation](https://arc.fuet.ch/concepts/modeling-bpmn/bpmn-2-0-notation-overview.md)

## Sources

- [Process mining - Wikipedia](https://en.wikipedia.org/wiki/Process_mining) — en.wikipedia.org, accessed 2026-11-09
- [What is Process Mining? Definition and benefits | Celonis](https://www.celonis.com/insights/topics/what-is-process-mining) — celonis.com, accessed 2026-11-09
- [IEEE 1849 - Wikipedia](https://en.wikipedia.org/wiki/IEEE_1849) — en.wikipedia.org, accessed 2026-11-09
- [Process Mining Manifesto - IEEE Task Force on Process Mining](https://www.tf-pm.org/resources/manifesto) — tf-pm.org, accessed 2026-11-09