Comparative Performance of Arm and Farm on a Normalised Datasets

Association rule mining is basically used to generate association rules on a real life datasets. A well-known algorithm called apriori is used to generate the frequent pattern itemsets for a given transaction. Since real life dataset consist of nominal, continuous, integer attribute fields, to convert it into binary format some type of pre-processing has to be done on the dataset. In this paper, we had evaluated the performance of two algorithms that is ARM(Association Rule Mining) and FARM(Fuzzy Association Rule Mining) on the basis of generation time by supplying different support and confidence values.for data pre-processing ,two methods are used: discretisation and normalisation. Discretisation converts the range of possible values of continuous data into subranges which is identified by a unique integer label .It also convert values associated with instances to corresponding integer label. Normalisation process converts values of nominal data into corresponding integer labels.