Research And Application On Information System Component Identification Method Based On Domain Modeling-oriented

With the rapid development of China's informatization construction,various large enterprises and institutions have successively built many application systems.In recent years,business operations and processes of enterprises have become more and more complicated.The demand for software is changing dramatically,but information systems cannot quickly adapt to the huge changes in external business processes.Increasing the efficiency of software reuse is the key to solving this problem.Under this background,Component-Based Software Development(CBSD)came into being.In fact,when external requirements change,based on CBSD,information system developers can quickly locate the components that need to be modified,which means that it can quickly respond to changes in external requirements,which is of great significance for improving software reuse and development quality.As the key technology of software reuse,component identification has always been a hot and difficult point of CBSD research.Through the appropriate business component identification model,under the premise that the component granularity is appropriate,the identified classes(objects)within each business component are closely connected,and at the same time,the interaction between various business components is reduced as much as possible,which makes each business process could almost implement completely within a business component.Therefore,the information system can quickly adapt to changes in external business processes,and can truly achieve so-called software reuse.This dissertation mainly analyzes the research status and related literature of information system component identification technology,and combines the actual situation and relevant theoretical and technical foundation to propose a complete information system component identification method based on domain-oriented modeling,which specifically includes FFCA-based information system component identification model and validation process of it based on graph entropy.Firstly,we could identify the fine-grained boundary,control and entity classes using the domain modeling.Based on FFCA,the quantitative method is used to calculate the consistency of create,delete,update and read operations of entity objects to get the identification of entity components;consider the user permissions of the boundary classes and the call between the boundary and control classes,refer to the CRUD guidelines,and use the quantitative calculation method to further integrate the boundary and control classes into the entity components.In the end,every business component contains closely related boundary,control and entity classes.Secondly,this dissertation determines that the degree of cohesion,coupling,and granularity suitability are the indicators for the validity test of business component identification results.On this basis,we use graph entropy theory to improve the traditional research,which researched the cohesion and coupling of information systems based on information entropy.At the same time we propose an improved method for measuring the cohesion and coupling of information system components,and a formula for calculating the suitability of business component granularity,which could further complete the validation process of information system component identification.Finally,through the complete business investigation and domain modeling for the material management business process of Beiran Tongzhou,we realize the design and development of material management system based on the component identification,and obtain practical material management industry universal business component library,which helps relevant companies in the industry to improve their own information management level and enhance their own industry competitiveness.