Incremental Mining of Association Rules: A Survey

The association rule mining has been very useful in many applications such as, market analysis, web data analysis, decision making, knowing customer trends etc. In transactional databases as time advances, new transactions are being added and obsolete transactions are discarded. Incremental mining deals with generating association rules based on available knowledge (obtained from mining of previously stored databases) and incremented databases only, without scanning the previously mined databases again. Several research works have been carried out for deriving the association rules and maintaining them efficiently without re-scanning the complete database. In this paper, a survey on different algorithms designed for incremental mining is presented. The algorithms are discussed into two sub-categories namely, apriori based algorithms and tree based algorithms. The pros and cons of these algorithms are also discussed in brief.