Research On Data Schema Evolution Technology Of Enterprise Data Space

With the advent of the era of big data,the enterprise data is becoming more and more complex and business needs are changing more and more rapidly.The existing relational database management mode can't deal with changes in business needs in a timely and rapid manner.When business needs change,the heterogeneous and multi-source data schema can't be modified in time to meet the new data application requirements and affect the normal operation of the data management system.It takes a lot of time,manpower and material resources to resolve the problems caused by business changes,the long cycle,high cost,and low efficiency method,can't meet the needs of rapid development of enterprises.In response to the above problems,this paper proposes the data schema evolution technology of enterprise data space,provides a set of related data schema evolution theory,finds the differences between data schema through data schema matching,and automatically calculates the operation sequence of data schema evolution to solve the challenges of data management brought about by frequent changes in business requirements.First,a hierarchical organization model is proposed to uniformly organize and manage various heterogeneous data of the enterprise,use the data resource catalog to flexibly organize data in multiple dimensions and multiple angles,and use the property graph model to uniformly describe various and management heterogeneous data schema.Secondly,based on the unified description and management of various heterogeneous data models using the property graph model,according to the structural information and element information of the data model,the initial similarity calculation method and the specific method of matching result filtering are changed to improve Similarity Flooding algorithm,through manual proofreading to get more accurate data schema matching results,to get the difference between data schema.Finally,the conflicts generated during the evolution of the data schema are classified,and strategies and specific solutions to resolve the conflicts in the data schema are proposed.The basic operations and combination operations of the data schema evolution are formulated,and the static calculation strategy of the data schema is used to automatically calculate the operation sequence of data schema evolution finally realizes the data schema evolution of enterprise data space.In summary,this paper presents the data schema evolution technology of the enterprise data space,designs and implements the data schema evolution system of the enterprise data space,and uses the real data in the enterprise to conduct the data schema evolution experiment,which verifies the effectiveness and feasibility of the research.This paperprovides a way for companies to better respond to the challenges caused by frequent changes in actual business needs,reduces the cost of enterprise data management,and improves the efficiency of enterprise data management.