Mining Time-Profiled Associations: An Extended Abstract

A time-profiled association is an association pattern consistent with a query sequence over time, e.g., identifying the interacting relationship of droughts and wild fires in Australia with the El Nino phenomenon in the past 50 years. Traditional association rule mining approaches reveal the generic dependency among variables in association patterns but do not capture the evolution of these patterns over time. Incorporating the temporal evolution of association patterns and identifying the co-occurring patterns consistent over time can be done by time-profiled association mining. Mining time-profiled associations is computationally challenging due to the large size of the itemset space and the long time points in practice. In this paper, we propose a novel one-step algorithm to unify the generation of statistical parameter sequences and sequence retrieval. The proposed algorithm substantially reduces the itemset search space by pruning candidate itemsets based on the monotone property of the lower bounding measure of the sequence of statistical parameters. Experimental results show that our algorithm outperforms a naive approach.