The k-modes algorithm presented in this paper is a simplification of the k-prototypes algorithm by only taking categorical attributes into account.// Therefore, weight g is no longer necessary in the algorithm because of the disappearance of SN. If numeric attributes are involved in a data set, we categories them using a method as described in (Anderberg 1973).// The biggest advantage of this algorithm is that it is scalable to very large data sets.// Tested with a health insurance data set consisting of half a million records and 34 categorical attributes, this algorithm has shown a capability of clustering the data set into 100 clusters in about a hour using a single processor of a Sun Enterprise 4000 computer.