Adaptations of data mining methodologies: a systematic literature review.

Area of Science:

• Data Science
• Information Systems
• Business Analytics

Background:

• Data mining methodologies (CRISP-DM, KDD, SEMMA) are widely adopted.
• Understanding practical application and adaptation of these methodologies remains limited.

Purpose of the Study:

• To investigate how data mining methodologies are used in practice.
• To identify the extent and nature of adaptations to standard data mining methodologies.

Main Methods:

• Systematic literature review of 207 peer-reviewed and grey publications.
• Analysis focused on the context and adaptations of data mining methodologies.

Main Results:

• Data mining methodologies are predominantly used 'as-is'.
• Adaptations are increasing, focusing on granular modifications and extensions.
• Key adaptation drivers include Big Data technologies (technological) and business process integration (organizational).

Conclusions:

• Standard methodologies often neglect crucial deployment issues for integrating data mining into IT and business processes.
• Future refinements should integrate data, technological, and organizational aspects to bridge these gaps.