ComputingComputer ScienceBusiness IntelligenceData MiningDatabasesFrequent Pattern MiningTemporal Data MiningHigh-utility Pattern Mining
High utility itemset (HUI) mining is a popular data mining task, which consists of discovering sets of items generating high profit in a transaction database. Recently, several efficient algorithms have been proposed for this task. But, most of them do not consider the on-shelf time periods of items, which thus lead to a bias toward items having more shelf time. Moreover, most algorithms cannot handle databases containing items with a negative unit profit, although this case is very common in real transaction databases. In this paper, we address both of these challenges by proposing a novel efficient algorithm named FOSHU (Faster On-Shelf High Utility itemset miner) to mine HUIs while considering on-shelf time periods of items, and items having positive and/or negative unit profit. An extensive experimental study with real-life datasets shows that the proposed algorithm can be up more than 1000 times faster and use up to 10 times less memory than the state-of-the-art algorithm TS-HOUN for this task. Moreover, experiments show that the proposed algorithm performs well on dense database and databases containing many time periods.
10
Mining high utility itemsets without candidate generation
Mengchi Liu, Junfeng Qu · 2012 · 647 citations
SPMF: a Java open-source pattern mining library
Philippe Fournier‐Viger, Antonio Gomariz, Ted Gueniche et al. · Journal of Machine Learning Research · 2014 · 417 citations