process mining

What is the purpose of discovery in process mining?

creates an as-is model from the event log, reflecting the actual process flow.

What does performance mining find?

identifies delays, bottlenecks, rework, and workload within processes.

What is the difference between task mining and process mining?

Task mining analyzes user actions at a desktop level, while process mining examines overall business processes using event logs.

What elements are part of a BPMN model?

tasks (work performed), events (something happens), sequence flows (arrows/orders), gateways (decisions/branches), and pools/lane (responsibility).

What issues can arise from missing or bad timestamps in an event log?

can cause incorrect order and duration calculations.

What is over-processing?

unnecessary work that does not add value, often occurring due to poor triage in case handling.

What is the fitness score in conformance checking?

indicates how well the event log aligns with the reference model; a high score means good conformity, while a low score indicates many deviations.

What are common violations identified in conformance checking?

skipped approvals, double payments, refunds before products are received, and wrong activity orders.

What is process mining?

uses event logs from IT systems to analyze how business processes actually happen.

What are the mandatory attributes of an event log?

Case ID, Activity Name, and Timestamp.

What is a bottleneck?

a slow point in the process that causes delays.

What is variant analysis?

compares groups of cases to identify differences in cycle time, flow, frequency, rework, bottlenecks, resources, and compliance.

What is conformance checking?

involves comparing the actual process recorded in the event log with the expected process model, business rules, or BPMN model. It helps identify deviations such as skipped steps, repeated steps, wrong activity orders, SLA breaches, and incorrect resource allocations.

What are the types of deviations identified in conformance checking?

• Flow conformance: Checks the correct activity order (e.g., approvals skipped before payment).

• Temporal/SLA conformance: Checks adherence to time rules (e.g., a case took 20 days when the SLA is 14 days).

• Resource conformance: Checks who performed activities (e.g., the same person created and approved an invoice).

• Exception analysis: Examines rare or unusual variants in case processing paths that are not commonly followed.

What tools does Apromore provide for conformance checking?

• Path filter: Checks the activity order or relation.

• Performance filter: Checks for long durations or SLA breaches.

• Rework filter: Identifies repeated activities.

• Case variant filter: Checks for unusual case variants.

• Frequency overlay: Visualizes skipped or repeated activities.

What are the main ethical issues in process mining?

• Privacy: Protecting personal information.

• Data minimisation: Collecting only the necessary data.

• Anonymisation: Hiding identities in datasets.

• Transparency: Explaining how data will be used.

• Consent: Ensuring individuals know and agree to data usage.

• Fairness: Avoiding biased analysis.

• Access control: Limiting data access to authorized individuals.

• Accountability: Assigning someone to be responsible for ethical data use.

What does Process Mining 1.0 focus on?

focuses on analyzing what has already happened in processes. It includes:

• Automated process discovery

• Performance mining

• Conformance checking

• Variant analysis

What is the goal of Process Mining 2.0?

predict and improve future process outcomes. It encompasses features such as predictive process monitoring, prescriptive process monitoring, robotic process mining, automated process improvement, and search-based process optimization.

What is predictive process monitoring?

Predictive process monitoring predicts future outcomes of running cases, such as the next activity, remaining time, final outcomes, or SLA breach risks. For example, it can predict that a claim is likely to be delayed by 5 days.

What is prescriptive process monitoring?

provides recommendations on actions to take to achieve better outcomes. For instance, it may suggest reassigning a case to another staff member to avoid SLA breaches.

What is robotic process mining?

identifies repetitive, rule-based tasks that can be automated using Robotic Process Automation (RPA). An example includes automating the task of copying customer details from emails into spreadsheets.

What are examples of automated process improvement?

uses data to suggest actionable changes to processes. Examples include:

• Removing unnecessary tasks

• Adding staff to bottleneck activities

• Automating a specific task

• Changing activity order or decision rules.