Visualization Oriented Data Model Quality Analysis System

With our country entering the era of big data,which is closely connected with people,things and networks,people from all walks of life tend to establish data model independently by combining big data with business requirements.However,the quality of the basic data used in the data model is the basis for making accurate decisions.If lacking accurate data,the result of display will be not convincing.Therefore,we must pay attention to the data quality problem.When we provide decision-making to users,we must be sure that the data used is safe and reliable.Based on the above background,this paper will design and implement a visualization oriented data model quality analysis system.The core of this system is to realize the quality management of data model.Firstly,the traceability information function is used to manage the relationship between data objects and provide technical support for model similarity analysis and index specification analysis.Secondly,we analyze similarity on the basis of correlation dimension of data model and normative dimension of indicators.Thirdly,the similar indicators can be merged into certain indicator.What's more,we evaluate the quality of data table on the basis of the change of data update.Finally,combining the data quality indicator from the analysis of data table with normalization indicator from the similarity analysis of index and similarity index of model to assess the quality of data model.The front-end mainly uses E-charts and Knowledge Graph for displaying data.The back-end uses Java language to program system.Spring Boot is used to build the framework.This system uses relational database named Mysql and graph database named HugeGraph.The system is mainly divided into five functional modules: metadata management module,traceability information management,model similarity analysis,index specification analysis and quality analysis.Finally,the system functions are displayed on the page and tested.Through the simple and intuitive visual interface design,the system displays the data and evaluate the reliability of the data source.The system uses the current hot topics such as machine learning,deep learning and combines the algorithm theory with the actual development of the system.