Frequent Patterns for Mining Association Rule in Improved Apriori Algorithm

An important aspect of data mining is to discover association rules among large number of item sets. Association rules are if/then statements that help uncover relationships between seemingly unrelated data in a relational database or other information repository. The main problem is the generation of candidate set. In this thesis we have presented a different algorithm for mining frequent patterns in large datasets using transposition and Boolean matrix of the database with modification of the Apriori-like algorithm. The main advantage of this method is that the database stores in different formats in each iteration. Database is filtered and reduced by generating the transaction ID for each pattern. In proposed method we describe the weighted Apriori algorithm, in that we collect the large amount of items divided and categorized into groups that is use of bitpartition technique to reduce the huge computing time and also to decrease the database size. Due to this the efficiency of algorithm increases. INDEX TERMS Data mining, association rule, candidate set, Apriori algorithm.