The knowledge discovery process

This chapter gives an overview of the knowledge discovery process. The full process starts from the definition and analysis of the business problem, followed by understanding and preparation of data, setup of the search for knowledge, the actual search, application of results in solving the business problem, and, finally, deployment and practical evaluation of the solutions. We outline the main tasks and methods that apply in each phase and we make references to the relevant chapters of this handbook. The discovery process is a combination of human involvement and autonomous methods of discovery. Autonomous methods may include automated task integration, for instance, integration of variable selection, knowledge mining, and result optimization. We also emphasize the iterative character of discovery including feedback loops and knowledge refinement.