Research On Incremental Updating Algorithm For Mining Association Rules

Data Mining as a technology for getting valuable information and knowledge from mass of data , theses days has been widely used . Association Rules Mining is an important branch of Data Mining , also is the first one of the issues . Apriori algorithm is a classical one in Association Rules Mining , many other algorithms are on the basis of it for improve-ment , and most algorithms are aimed at a static database .In practical applications , the data of the databases often needs to change , if use the previous algorithm to mine the rules , it will re-run the algorithm to the updated database, so the efficiency will be very low . Consequently , the Incremental Updating Algorithm for mining Association Rules appeared . There are two typical examples : FUP algorithm and IUA algorithm in the Incremental Updating Algorithm for Minning Association Rules . In this paper , we just discuss FUP algorithm . After analyzing and studying the FUP algorithm's defect , We proposed a new Incremental Update algorithm: TMFUP algorithm. The data format in the algorithm is vertical , compared with FUP algorithm , TMFUP algorithm can significantly decrease the number of scaning the original database , and the algorithm's implementation is more efficient.