A data mining tool for producing characteristic classifications in the legal domain

Data mining is a relatively new term for the discovery of knowledge in large data sets. Essentially the idea is to extract relationships from a mass of data. We suggest that opportunities for data mining exist in the legal domain, and we begin by suggesting some of the uses to which it might be put. We then briefly discuss some of the constraints that must be placed on a data mining tool in the legal domain. The bulk of the paper is concerned with a particular data mining tool, which is designed to identify the characteristics of typical members of a legal category. We describe the theory of the tool, and discuss some potential pitfalls with respect to its application. We then work through a simple example to illustrate how the tool operates