Yoğun nesne kümelerinin güncellenmesi için sıralamaya dayalı artımlı bir yaklaşım

ABSTRACT A CONSTRAINT-BASED INCREMENTAL APPROACH FOR UPDATE OF LARGE ITEMSETS Engin Demir M.S. in Computer Engineering Supervisor: Prof. Dr. Erol Arkun August, 2001 The main focus of data mining, subject of numerous studies in recent years, is to extract and analyze hidden knowledge from the collected and stored massive amounts of data. In order to manage this process, many automated knowledge discovery techniques have been developed. Discovery of association rules, which is the process of mining interesting and frequent patterns from data, is an important class of data mining. Most of the researchers proposed techniques for static databases but in real applications dynamic approaches are necessary to maintain the frequent patterns in terms of association rule mining. Additionally, from the standpoint of users, previously developed systems have lack of user exploration and control, lack of focus, and rigid notion of relationships. In this thesis, we propose a solution to both of the shortcomings of the previous systems. We integrate the anti-monotonicity and succinctness constraints to the previously developed update with early pruning (UWEP) algorithm. Keywords: Data mining, constraint based, association rules, update of large item- sets, pruning. m