RCRDE: A Method for Reducing the Rate of Re-Clustering, using Replicated Data Eliminate Algorithm

In this paper is explored a way to reduce therate of re-clustering andspeed uptheclusteringprocess oncategoricaltime-evolving data. This method introducestwoalgorithmsRDE (Replicated Data Elimination) andRCRDE. The RDEalgorithmremoves the successivesurveysof replicated dataandconsiders counters tokeepthis data. Hence the number of created windows via thesliding window techniqueis limited and thisleads todecrease thenumber ofimplementations ofclusteringalgorithm. The RCRDEalgorithmbased on MARDL (MAximal Resemblance Data Labeling) framework decidesabout re-clustering implementation ormodificationofpreviousclusteringresults. Thepresentedmethodisindependent of clusteringalgorithm'stype and any kind ofcategoricalclusteringalgorithmcan be used. According tothe results obtainedondifferentdata sets,this method performs well in practice and facilitatestheclustering implementationon categorical data. Also, this method can be utilized to cluster a very large categorical static databasewith higher quality than previous work.