Association Rule Mining by Discretization of Agricultural Data Using Extended Partitioning Algorithm

The agricultural dataset is a group of interrelated, discrete items of data that may be used for numerous data groups such as rural development project data, soil data, production advice and productivity data, land use data, pest and disease management data, etc. Aggregation of multidimensional attributes of agricultural data for mining quantitative association rules requires additional scans of large databases. It reduces the performance of the Partitioning Algorithm (PA) used for mining quantitative association rules. Extended Partitioning Algorithm (EPA) has been implemented in the present paper to reduce the number of additional scans for aggregation of multidimensional attributes at different levels. The number of association rules mined from EPA has been compared with the number of association rules mined from PA.