Set Representation for Itemsets in Association Rule Mining
Carynthia Kharkongor, Bhabesh Nath · 2018
Frequent itemset mining is one of the main problems in data mining. Many mining algorithms have been proposed to give the solution. Handling small database is easier than compared to large ones. However, when large databases are concerned, the problem of storing the itemsets in the memory and the time consumed for executing the algorithm arise. The frequent and candidate itemsets do not fit in the memory and have to be brought back into the main memory for computation. This involves disk I/O operations that increase the cost and as well as wastage of CPU time. Representation of itemsets plays an important role in consideration with the memory consumption. In this paper, representation of itemsets using bitmap is introduced for association rule mining that will reduce the memory consumption and saves the execution time. This will eventually improve the process of mining the itemsets.