Relational data describes objects by specifying pairwise similarities
or dissimilarities between objects, while object data describes the ob
jects through feature measurements. This paper shows how c-means clust
ering methodology for object data can be applied to relational data.