Immune Optimization Based Genetic Algorithm for Incremental Association Rules Mining

Business activity and engineering practice always produce large data sets carrying important information, but because of the data sets´ largeness and frequent updating, if we apply the Apriori based algorithms to them for incremental rules mining, it is not only inefficient, but also either redundant rules would be produced under low threshold of minimal support, which makes users hardly distinguish which rules are really meaningful, or significant rules with low support in additional data set would possibly lost when the threshold is defined high. Motivated by these, therefore, following genetic principles, and combining with natural immune evolution theory and relevant bionic mechanism, this paper proposes an IOGA (immune optimization based genetic algorithm) approach for incremental association rules mining to large and frequent updating data sets. Experiment demonstrates the method´s efficiency and presents its good performance in pruning redundant rules and discovering meaningful rules, perceiving low support rules in additional data set.