Lecture 3 - Process mining

Question 1

What is a process?

Answer 1

Sequence of steps to accomplish some goal.

Question 2

What is process mining? (4)

Answer 2

1. Extracting knowledge from event logs; data about the events in the process.
2. Building a process model inductively; look at different cases and therefrom learn what is the process structure.
3. Comparing mined process model with process description
4. Visualizing mined models for analysis

Question 3

Process mining vs Data mining

Answer 3

1. Data mining is the computing process of discovering PATTERNS in large data sets.
2. Process mining aims at discovering the PROCESS, works with event and relationships between events, no numerical variables.
3. Commonalities:
   - inductive algorithms
   - overfitting
   - visualization

Question 4

Three main types of process mining (vd Aalst):

Answer 4

1. Discovery: discover the actual process in the business
2. Conformance: compare process model with a standard model (process description or standard)
3. Enhancement: try to improve the process model

Question 5

Quality criteria of a discovered model (4):

Answer 5

1. Fitness
2. Precision
3. Generalization
4. Simplicity