Affiliations: School of Economics and Management, Chongqing Jiaotong University, Chongqing, China
Correspondence:
[*]
Corresponding author. Bing Xu, School of Economics and Management, Chongqing Jiaotong University, Chongqing, China. E-mail: xbing@cqjut.edu.cn.
Note: [1] Bing Xu (1966-), female, MBA, Lecturer of School of Economics and Management, Chongqing Jiaotong University, mainly engaged in teaching and research on international economy and trade.
Abstract: In the process of e-commerce transactions, a large amount of data will be generated, whose effective classification is one of current research hotspots. An improved feature selection method was proposed based on the characteristics of Bayesian classification algorithm. Due to the long training and testing time of modern large-scale data classification on a single computer, a data classification algorithm based on Naive Bayes was designed and implemented on the Hadoop distributed platform. The experimental results showed that the improved algorithm could effectively improve the accuracy of classification, and the designed parallel Bayesian data classification algorithm had high efficiency, which was suitable for the processing and analysis of massive data.
