Mining Algorithm Of Frequent Items Based On Item Adjacentcy List And Trasaction Tree

Data mining technology is an effective approach to resolve the problems of abundant data and scanty information. It currently is the research frontier within the information science filed. It has deemed to a field that has broad prospect of application in database research. In many data mining algorithms, mining association rule is an important matter in data mining, in which process that mining frequent itemsets is a key problem in mining association rule. Apriori algorithm, FP-growth algorithm and Eclat algorithm are two classical frequent itemsets mining algorithms. Though they improved the efficiency of the algorithm in a certain extent, they exisited some problems, such as the produced candidate itemsets, the times of scanning database and the request of memory. On the base of analysis of the classical algorithms, we research the problem of mining association rule based on matrix in the following aspects.Firstly, we find the problems of existing frequent itemsets mining algorithms and of frequent itemsets mining through analyzing them, which produce many candidate itemsets and scan database many times. Aiming at these problems, with the theory of vector and item adjacency list, arrange items from transaction datebase into item adjacency list.Secondly, For the shortage of existing algorithm in find of frequent items suah as numerous search-designate database set too many times and gennerate too much candidate itemsets. Range items from transaction datebase into transaction tree.Using the transaction tree can quickly find the maximal frequent itemsets.Finally, There exisits many redundance and invalid association rules, or neglects many usable association rules in traditional association rule mining algorithms. We bring forward an effective association rule mining algorithm, in which we add effective confidence, a new measure standard. Effective association rules can be produced in this algorithm.