Process overview
CRISP-DM is a loop because reality pushes back.
The five-minute version
CRISP-DM is the Cross-Industry Standard Process for Data Mining. Its six phases are not a waterfall checklist. They are a practical way to keep a data project tied to a decision, grounded in the available evidence, and honest when the first answer is not good enough.
Good teams move forward and backward through the phases. A model can expose a data problem. An evaluation can reveal that the original business question was too vague. Deployment can prove that the world changed after the model was trained.
- 1 Business Understanding Turn a vague business problem into a precise question.
- 2 Data Understanding Check what data exists, what it means, and where it is weak.
- 3 Data Preparation Turn scattered raw records into a dataset a model can trust.
- 4 Modeling Use patterns in the prepared data to make a useful prediction.
- 5 Evaluation Test whether the result is good enough for the decision it supports.
- 6 Deployment Put the solution into use, monitor it, and improve it when reality shifts.