Data Analysis And Forecast In The Field Of Marketing

On the basis of the data of customer's consumption behavior,this thesis carries out in-depth research on the problems of Association Rules Mining for customer behavior,customer value analysis and prediction,and marketing platform data security,which provides support for enterprises to create efficient and secure new marketing mode.The main research contents are as follows:(1)In order to solve the problem of customer behavior association rules mining,the interest model is used to improve the Apriori algorithm,and then MapReduce algorithm,so that it can be run on Hadoop cloud computing platform for massive data mining.(2)To accurately predict the customer value,with the means of RFM model,Markov chain properties,NBD distribution model,Gamma-Gamma mixture model and K-means clustering analysis algorithm,build customer value analysis of stochastic models,then introduce the naive Bayes to build customer value classification and prediction model,which effectively analyzes the customer value.(3)Regarding data security of mobile marketing platform,to improve the operation efficiency of RSA algorithm,firstly combination of Carb Monte algorithm with the Miller-Rabin prime test optimization strategy,fast random strong Prime algorithm is designed to improve the efficiency of initialization.Secondly,the MMRC algorithm is used to optimize the RSA decryption process.Then the M-ary algorithm is introduced to optimize the operation of the modular power in the process.Finally conduct coding in the Android module.At last,this paper describes the structure design of mobile marketing management system and the design of Hadoop cloud platform.