Architectural perspective of parallelizing association rule mining

Association rule mining algorithms play a major role in the data mining research. Association rule mining algorithm is one of the data mining algorithms used to find the association between the items in the item set. Association rule mining is not only used widely for finding the patterns, it can also be extended for prediction. Its major drawback is that, it consumes a computational time. In order to reduce the execution time, the parallel and distributed based algorithms are brought into play. Hence in this paper, we highlight the architectural aspects related to parallel association rule mining (PARM). The ways of achieving parallelism and in turn reduced computational time by employing task and data parallelism in multiprocessing and multicomputing environment is analyzed in this paper. Another major issue encountered by PARM is load balancing. To overcome this, we suggest the incorporation of dynamic load balancing in PARM. Grid computing can be used both for compute intensive tasks and data intensive applications as they proffer resources and data access mechanisms. Grid can offer a computing and data management infrastructure for supporting decentralized and parallel data analysis. In the grid environment, mining system developers can focus on the mining applications and not the issues associated with distributed processing. In this paper, we have also given the infrastructure needed for grid based distributed association mining.