doi:10.3850/978-981-08-7300-4_0352

ABSTRACT

The real transactional databases often show evidence of temporal characteristic and time varying behavior. Given a time-stamped transaction database, time-profiled associations represent those subsets of items whose support time sequence satisfies a given specification. This paper proposes efficient similarity profiled temporal association mining using FP-tree data structures. Efficiency of FP-SPAMINE is achieved with three characteristics: (1) a large database is compressed into a condensed, smaller data structure, which avoids costly, repeated database scans, (2) FP-tree based multi level mining adopts a pattern fragment growth method to avoid the costly generation of a large number of candidate sets and (3) a partitioning based, divide–and–conquer method is used to decompose the mining task into a set of smaller tasks for mining are confined patterns in conditional databases, which dramatically reduces the search space. This approach is implemented on a synthetic dataset and result shows that FP-SPAMINE gives better performance over an SPAMINE approach.

Keywords: Temporal data mining, Temporal association pattern, Similarity, FP-tree.

© 2011 ICCA 2010 Organizers