The Parallel Association Rules Algorithm Based On Mapreduce In The Application Of Community Analysis Research

Due to rapidly development of the information technology and furtherpopularization of the Internet application, this situation lead to the result that the datarevealed a trend of quantitative sea. Facing the increasingly data of large community,if we can excavate potential information and knowledge from the vast amount of dataseeming complex and no chapter. It is the right way to make the enterprise occupy thedominant position in the fierce competition firmly. Association rules algorithm iswidely used in the community network of potential information mining in recent years.However the time of analysis and calculation is greatly increased because of the dataevolution of quantitative sea, the traditional association rules mining algorithm hasbeen unable to work as efficient as the original data including processing,performance and operation efficiency.The development of cloud computing technology bring in a new technologyinnovation to association rules algorithm. Cloud computing platform, with its strongexpansibility in huge amounts of data processing, can add storage dynamically. If wecan optimize the association rule mining algorithm for cloud model and deploy tocloud computing platforms, certainly a good way to solve mass data mining problemscan be found in the community.This paper make an introduce of the concept of cloud computing and relatedtechnologies firstly and then focus on further description of MapReduce programmingmodel, calculation process, and the research significance. In view of disadvantages ofthe traditional Apriori algorithm of association rules in the community, this paperapply to the algorithm of analysis and optimization, combining with the powerfuladvantages of Hadoop platform, and implement the AprioriTid-Partition algorithm.The algorithm is based on the framework of MapReduce, realizing parallel computingby Map and Reduce model. In the end, this paper carried on the contrast experimentto test the accelerate ratio and applied to the community in data analysis. By theexperimental verification, the algorithm of huge amounts of data in communityprocessing efficiency has improved obviously compared with the traditionalassociation rules algorithm, and showed a good scalability and speeding ratio.