Application Research Of Data Mining Technology In Mobile Service Development Quality Diagnosis

In the face of the wave of the new generation of mobile Internet,telecom urgently needs a kind of data mining and analysis,which can help telecom business decision-making and help the development of enterprises.The development of quality diagnosis is an important guarantee for an enterprise to make profits,and also the key to win the competition.By establishing a set of large data wide tables,improving the business data reserve,and creating a new data analysis and processing method,the end-to-end data analysis and diagnosis process is realized.The whole process of data analysis is visualized,controllable and reversible,providing a more comprehensive and in-depth support for management decision-making Reliable analysis basis.Data analysis introduces modeling tools and modeling concepts,expands the depth and breadth of analysis,and provides core competitiveness for enterprises to achieve differentiated development in the future.This thesis analyzes the application of data mining in the quality diagnosis of mobile business development,and introduces the objects,functions and methods of data mining;two main data mining technologies of association rules and decision tree;and SPSS Clementine data mining tools.The main work of the thesis is as follows:This thesis analyzes the demand of mobile service development quality diagnosis system,uses the method of data mining to accurately find the quality characteristics of mobile users,and implements the TOP evaluation index of system monitoring link.The design of development quality diagnosis and analysis system includes three parts:data input,diagnosis model,output diagnosis report and problem list.Before the construction of the system,the data of the model is prepared through the database.Mobile development quality diagnosis system through wide table design,data preparation,sample selection,data validation to carry out the system process design,and through the analysis period of different data samples,off-line caliber,payment method and through factor analysis to screen the modeling input variables.Finally,through the comparison between C5.0 decision tree and binomial logistic regression,it is concluded that C5.0 decision tree is used Data mining is a relatively stable off network prediction model.Finally,establish and improve the function design and analysis tree configuration of the system for mobile business development quality diagnosis,realize the visualization,controllability and reversibility of mobile service development quality diagnosis,and provide strong guarantee for targeted marketing and differentiated development...