Research on distributed mining algorithm for association rules oriented mass data

Considering the cost, safety and competitive of data migration, a distributed association rule mining algorithm based on matrix named DARMO is put forward for some special distributed applications. This algorithm has some characteristics such as high degree of parallelism, fewer database scanning, less communication overhead and low complexity. The correctness of the algorithm is proved theoretically through the use of a completely different way against the classic Apriori algorithm to generate frequent item sets and avoiding cumbersome connections and pruning operations, and efficiency of the algorithm is improved. Finally, the complexity, parallel price, speedup and scalability of the algorithm is analyzed, and effectiveness of the algorithm is verified by example analysis and experimental simulation.