Towards An Incremental Maintenance of Cyclic Association Rules

Recently, the cyclic association rules have been introduced in order to discover rules from itemscharacterized by their regular variation over time. In real life situations, temporal databases are oftenappended or updated. Rescanning the whole database every time is highly expensive while existingincremental mining techniques can efficiently solve such a problem. In this paper, we propose anincremental algorithm for cyclic association rules maintenance. The carried out experiments of ourproposal stress on its efficiency and performance.