Categorical data clustering (CDC) and link clustering (LC) have been considered as separate research and application areas. The main focus of this paper is to investigate the commonalities between these two problems and the uses of these commonalities for the creation of new clustering algorithms for categorical data based on cross-fertilization between the two disjoint research fields. More precisely, we formally transform the CDC problem into an LC problem, and apply LC approach for clustering categorical data. Experimental results on real datasets show that LC based clustering method is competitive with existing CDC algorithms with respect to clustering accuracy.