What is CRISP DM?

CRISP-DM is created by IBM, stands for Cross-Industry Standard Process for Data Mining, is an industry proven way for data mining. Data Mining to insights process is CRISP DM.

Extracted from: https://www.ibm.com/docs/it/spss-modeler/saas?topic=dm-crisp-help-overview



The CRISP DM’s 6 data mining process steps are Business Understanding, Data Understanding, Data Preparation, Modeling, Evaluation, and Deployment. The arrows indicates frequent dependencies between the steps, but the sequence of the steps is not strict and can move back and forth between steps when necessary. Hence, you can change from “Data Understanding -> Data Preparation” to “Data Preparation -> Data Understanding”, meaning we do data cleaning first then analyze the data.



Business Understanding Step

Before working on the CRISP DM, you should explore what your organization expect to gain from the data mining project. Try to involve as many key people as possible and the final step is to produce a project plan.

- Determine Business Objectives: Describe the problem you want to solve, specify all business questions as precise as possible, Determine any other business requirements, specify expected benefits in business

- Assessing the situations

- Determining Data Mining Goals…