and P.Ronkainen; Similarity of Attributes by External Probes

In data mining, similarity or distance between attrib-utes is one of the central notions. Such a notion can be used to build attribute hierarchies etc. Similarity metrics can be user-defined, but an important prob-lem is defining similarity on the basis of data. Several methods based on statistical techniques exist. For de-fining the similarity between two attributes A and B they typically consider only the values of A and B, not the other attributes. We describe how a similarity no-tion between attributes can be defined by considering the values of other attributes. The basic idea is that in a 0/1 relation r, two attributes A and B are similar if the subrelations aA=l(r) and aB=~(r) are similar. Similarity between the two relations is defined by con-sidering the marginal frequencies of a selected subset of other attributes. We show that the framework pro-duces natural notions of similarity. Empirical results on the Reuters-21578 document dataset show, for ex-ample, how natural classifications for countries can be discovered from keyword distributions in documents. The similarity notion is easily computable with scal-able algorithms