| Summary: | In practice, single item support cannot comprehensively address the complexity of items in large datasets. In this study, we propose a big data analytics framework (named Multiple Item Support Frequent Patterns, MISFP-growth algorithm) that uses Hadoop-based parallel computing to achieve high-efficiency mining of itemsets with multiple item supports (MIS). The proposed architecture consists of two phases. First, in the counting support phase, a Hadoop MapReduce architecture is employed to determine the support for each item. Next, in the analytics phase, sub-transaction blocks are generated according to MIS and the MISFP-growth algorithm identifies the frequency of patterns. To facilitate decision makers in setting MIS, we also propose the concept of classification of item (COI), which classifies items of higher homogeneity into the same class, by which the items inherit class support as their item support. Three experiments were implemented to validate the proposed Hadoop-based MISFP-growth algorithm. The experimental results show approximately 38% reduction in the execution time on parallel architectures. The proposed MISFP-growth algorithm can be implemented on the distributed computing framework. Furthermore, according to the experimental results, the enhanced performance of the proposed algorithm indicates that it could have big data analytics applications.
|
|---|