Design Of Group Insurance Data Analysis System In Large Data Age

With the development of computer systems, insurancecompanies require in-depth analysis of the data and application.A lot of domestic insurance companies start to develop their owndata analysis system, but it always has very high failure rate thata lot of projects are unfinished or stopping use them atfer6months or1year.The sotfware requirements for the earlier period are informationtechnology with clear demand and implementation steps, goodcommunication with program design staff. Different fromtraditional sotfware design, data analysis is a process ofknowledge discovery that th’e users goal is to get a lot of usefulknowledge from data, but it is a little hard for them to describe itclearly. Sometimes this process is interactive.In recent years, companies continue to launch a lot of newbusiness models with typical internet character. These newbusiness models are pushing the insurance companies to make agreat changes that is changing the traditional way ofdecision-making and operations and thus enhances the customerexperience and operational efficiency. Big Data thought andtechnology may be a good way to solve the current problems ofthe insurance companies’ data analysisIn order to design a effective data analysis procedures forinsurance enterprises, the paper carries out the following work:summarizes the current status of the development of large datathought, mainstream thinking and current limitations;summarized the general practices and their pros and cons for theprogram design; solves the information silos of data analysisproblems, the program scalability issues, long cycle fordevelopment, and so on.Data analysis program designed is to change the past userdemand-driven model, no longer passively accept the needs of users during the design process, but more concerned about therelationship between information users. Using multi-platform,multi-source data source collection methods, and effectivesolution to clean up the problem of multiple data sources andintegration, we could build a unified data analysis services forthe multi-dimensional data marts.Through continuous improvement in work practices，the currentdata analysis program can be a good decision support services inthe enterprises, the policy rate payment estimates, analysis andoperational quality of early warning.