Design And Implementation Of Regional Production Data Processing System Of MCC CISDI

The development of production enterprises will always be accompanied by problems,and the era of massive data increases not only brings business opportunities,but also a severe test.In an environment where the data presents 4V(large,diverse,high-speed,low-value density)signs,how to properly use the data obtained from various aspects of the enterprise is the problem.In this regard,after de-noising,de-duplication,and establishing association rules for the data,mining data is used to analyze the cleaned data,and visual methods are used to show users their value trends,so as to make full use of data value to achieve enterprise-level decision support and promote enterprise development.The regional production data processing system studied in this paper relies on the actual development project of MCC CISDI in the regional production,and aims to solve the problems of internal business data,real-time data,and offline storage of huge but redundant data of MCC CISDI.Use open source data mining tools to collect,preprocess,analyze,and mine multiple data sources to help guide production,management,and operations.The whole system is divided into data acquisition layer,data preprocessing layer,analysis and mining layer and visual display layer from bottom to top.This paper mainly completed the following three aspects of work:(1)By studying the principles of data mining,data mining types,mining processes,data mining prototypes,and data mining algorithms,combined with the analysis of the needs of MCC CISDI,a data mining mechanism was established for the massive data within the company,which laid a good foundation for the development of subsequent systems.(2)Through an overview of the topic selection background,current status of the field,and related theories,algorithms,and tools involved in completing the system,the following technical route is established: selecting the Springboot+Spring MVC+Mybatis framework set to build the overall system architecture;using the API at the application layer to call KNIME for data mining processing analysis,and use workflow(workflow model)to visualize the data mining algorithm as a model;use the relatively mature ECharts as a visualization tool,to show mining data results in the form of line charts,bar charts,scatter charts,etc.(3)In order to effectively use and analyze the data by using mining processing technology,the user needs are deeply analyzed,and the decision tree algorithm is used as an example to complete the design and implementation of the regional production data processing system.The design of the system structure,function and mining model are elaborated in detail.Mainly discusses the functions of reading data,separating data,marking classification,establishing decision tree,predicting classification,and evaluating decision tree in data mining applications,and using PMML(Predictive Model Markup Language)for the relevant performance of decision tree in the implementation process.The original model has been improved.The results show that the improved model can more effectively deal with data classification and model transplantation problems in the regional production data processing system.