Ultimate

Abstract

Listen

Translate article

Discovery of temporal association patterns from temporal databases is extensively studied by academic research community and applied in various industrial applications. Temporal association pattern discovery is extended to similarity based temporal association pattern discovery from time-stamped transaction databases by Sashi Sekhar, Yoo using methods to estimate support limits and distance limits of temporal item sets. Yoo also introduced methods for pruning using distance bounds and proposed SEQUENTIAL, SPAMINE approaches. Our previous research introduced algorithms G-SPAMINE, MASTER, Z-SPAMINE for time profiled association pattern discovery that applied distance measures SRIHASS, ASTRA, GARUDA, and KRISHNA SUDARSANA for similarity computations. SEQUENTIAL, SPAMINE, G-SPAMINE, MASTER, Z-SPAMINE approaches are all based on snapshot and lattice database scan strategies that prune itemsets making use of estimated support and distance values. The major limitation of all these algorithms is their inevitability to eliminate database scanning process for knowing true supports of itemsets. To eliminate the requirement of retaining database in main memory in order to know true supports, VRKSHA and GANDIVA are two pioneering research that introduced tree structure for time profiled association mining. VRKSHA is based on snapshot tree scan technique while GANDIVA is lattice tree scan based approach. VRKSHA and GANDIVA both apply Euclidean distance function but does not estimate support and distance bounds. This research introduced the pioneering work ULTIMATE that uses a novel tree structure generated using similarity measure ASTRA and applies support and distance bound computations for pruning temporal patterns. Experiment results showed that ULTIMATE outperforms SEQUENTIAL, SPAMINE, G-SPAMINE, MASTER, VRKSHA, GANDIVA algorithms.