Tom Khabaza: Abstract

Data mining provides benefits in a very wide variety of applications, and the process is well documented in the industry standard methodology CRISP-DM. However, although CRISP-DM describes the process accurately, it does not explain why data mining takes the form that it does. The 9 Laws of Data Mining are statements which can be observed to be true and are well-known to practising data miners; I will present these laws and propose explanations as to why they must be true. These explanations range from the definitional to the (apparently) impossible, and rest on theories about the nature of data mining.

Organised by BCS SGAI The Specialist Group on Artificial Intelligence http://www.bcs-sgai.org