Classification of customer requirements on Map Reduce-based Naive Bayes

In the era of continuously enlarging mass customization based on customer requirements, a Classification method called MP-NB (i.e. Map Reduce-based Naive Bayes) is proposed to process CRIA(i.e. Customer Requirement Information Acquisition) classifications on large-scale mobile data. It utilizes Hadoop2.X-based system to store CRIA on HDFS. Combination with the standardization theory of CRIA, It utilizes Map Reduce-based distributed computing framework to accomplish Parallelization Naive Bayes. According to the processing level, Extensive experiments on gigabyte of realistic mobile requirement information are classified on Hadoop. And the results show that MP-NB is characterized by superior efficiency and good scalability.