Research on Mining Association Rules in Distributed System

With the development of Internet and the distributed database technology, a great deal of data is stored in the disrtibuted web nodes and it is impossible to be stored in one single node on account of communication, efficiency and security. So it's a very important research in the data mining domain. This paper makes a thorough research on mining association rules in the distributed database system. It extends the most classical algorithm Apriori based on distributed transactional database system. The system is realized based on the local global association rules mining solution. Finally the paper puts forward an efficient algorithm of getting association rules from frequent itemsets. This improved algorithm shows sound extension, short time complexity, small communication cost and simplicity.