Visual Analysis And Mining With Professional Title Data

Data visualization and data mining is an important part in the informationmanagement system. The main research content of this paper is how to dig valuabledata from the vast title declaration data, and transform it into intuitive, graphicdisplayed, spatiotemporal changing visualize data to the manager.In the first, the paper studies the basic theory and method of data visualization andlearns and masters the foundation of related data visualization software and tools,aiming at data visualization business requirements in title assess informationmanagement system. The paper achieves the comparative analysis, regional analysis,data flow analysis, professional structure application analysis and evaluation resultsanalysis in title assess information management system by Circos and R languages.Meanwhile, according to the demand of managers to understand the requirementsof possible assign points with experts, the paper conducts the declaration data, scoreand results in the past title assess. The paper uses the multivariate linear regressionmodel and regression tree model to establish expert quantification natural endowmentsprediction model; the two models have been conducted by combined and contrastexperiments. The experiments show that the quantification natural endowmentsprediction model with combined expert has better prediction performance, which canevaluate title applicant primarily in a certain extent. Under the condition of thelimitation of title indicators, our work can select title applicants by the simulation ofprediction model, and provides the scientific basis for the title management departmentto fully grasp title related data and allocation the index of title this year, thus improvesthe working efficiency.