Research on Data Mining Algorithm Based on Cloud Platform

Yingxue Cai, Jia Chen, Song Hu, Hui Hu, Sibo Huang, Zhaoquan Cai*

Huizhou University, Huizhou, 516007, China

Abstract: In order to improve the parallel processing capability of cloud computing, it is necessary to mine data on cloud platform to improve the ability of data scheduling and recognition. A data mining algorithm is proposed based on feature extraction of closed frequent itemsets under cloud platform, designs the general framework of parallel closed frequent itemsets data mining and constructs a data flow model, and establishes a universal knowledge cloud model of Cloud-P2P. Big data in the cloud platform is pushed to the appropriate server through the information supply module, and the information is updated by the user node and the server to extract the closed frequent item characteristic quantity of the data information flow. Using semantic ontology fusion and association rule scheduling method, big data mining under cloud platform is realized. The simulation results show that the method has good output recall, high precision and strong anti-jamming ability in the data mining process.

Keywords: cloud platform; data mining Cloud-P2P cloud model; feature extraction