Inductive learning of characteristic concept descriptions from small sets of classified examples

Abstract

This paper presents a novel idea to the problem of learning concept descriptions from examples. Whereas most existing approaches rely on a large number of classified examples, the approach presented in the paper is aimed at being applicable when only a few examples are classified as positive (and negative) instances of a concept. The approach tries to take advantage of the information which can be induced from descriptions of unclassified objects using a conceptual clustering algorithm. The system Cola is described and results of applying Cola in two real-world domains are presented.