System Analysis & Design And Its Applications Based On Formal Concept Analysis

With the rapid development of computer technology and the urgently demand of informationization in various fields, software engineering projects become increasingly complex and large. The requirements of systems analysis, design, development, operation and maintenance, upgrading are unceasingly increasing. Take insurance industry as an example, as the domestic insurance industry growing fast, insurance companies pay much more attention to the enterprise informationization. Using information technology to support business activities is the way to maintain the competitive edge and better respond to international and domestic challenges brought by competition. Therefore, enterprises should have agile and flexible systems, but that often depends on the understanding and definition of the system of the system development process, that is, the traditional system analysis and design process. However, the large number of business data and the wide range of business kinds decide the complexity of the system. In such a complex industry, information systems need to be flexible. So that, we need a good method of analysis and design to support the definition of the system concepts and the reduction of the coupling among system elements, then we can make the system more flexible and easy to extend.To be highly refined, full decoupling of the business entity model and to resolve the system analysis problems of the traditional business model in an abstract non-uniform grain size and subjective factors interference, this paper explores an approach of system analysis and design that based on Formal Concept Analysis (FCA). Using the theories and methods of FCA for reference, we discuss about the business entities analysis and database design, module analysis and functional design, analysis and the role of authority, SOA reusable service components and system integration. And, we make a comprehensive analysis of the steps based on the method of analysis and design models. Based on a practical insurance system, we prove that the system analysis and design based on FCA works well in the process of business entity analysis, database analysis and database structure design.FCA is a branch of applied mathematics, which can classify the elements that being in some certain relationship or sharing common characters, and can find the concepts and the relationship between them, using mathematics way to express concepts and relationships. Based on an actual core business system of insurance, this paper chose some policies of different risk kinds for study, applied FCA to entities abstract, conceptual level analysis, clustering concept lattice properties, analysis of the database structure, and finally got the business entity database model. Practice has proved that the FCA-based method of analysis and design system in the insurance business entities analysis is feasible and effective.In this paper, the main work and innovation are:Summarizing the step of using the system analysis and design method that based on FCA: Choosing the right point for study, mark the "formal object" and "formal attribute"; to build the formal context base on the elements that have been selected; check semantic meaning and clear the context; to generate the corresponding Hasse diagram; Analyze the concepts and their relations; to modify the formal context; iterate the above steps until getting the satisfactory results.Apply FCA methods to analysis a new system. At present, in the field of software engineering, the applications of formal concept analysis in systems analysis and design are often based on the legacy systems or legacy databases. This article gets the business entities by choosing the elements from the carrier of information - the policy, and then using FCA-based method of analysis and design to analyze the Hasse diagram. Take these steps as a new way to abstract business entities and design database model of a new system.In the process of modifying formal context, we propose a way to examine the semantics meaning via industry-stand issues. To most system analysts, the understanding of the business is necessary. But the entire analysis and design process is often based on experience, training or knowledge that they have had. Now based on the industry-standard, analysis and design process can less rely on professional knowledge, as far as possible to reduce the cost of training and communication.The results of concept lattice analysis can be counterproductive to the policies. Analysis of the concept of policy is actually a high level of abstraction. If an abstract on the results shows that the policy is too fragmented, it means maybe there are certain problems in the policy design. The interaction between them can make the design of policy become standardizationThe application of system analysis and design base on FCA in new system analysis and design is an innovative attempt. Its visualization of the system concepts and their relationships can help developers understand the whole system more consistently and more intuitively, upon that, a more objective manner to design and a better architecture can be obtained. The loosely coupled concept hierarchy allows the system to be more flexible and extendable, and system components reusability are fully improve. Our study shows that the system analysis and design base on FCA breaks the traditional boundaries of the design process for system analysis. It provides a new abstract model of the system concepts in an effective way. For model validation and realization it provides a powerful support. The use of such method can obtain a highly abstract, decoupling and objective system model. This formal analysis method and model make up a great understanding of the differences among businesses and reduce the human interference on the system. It raises the software quality control to the analysis and design level. Even though the specific application of the system analysis and design base on FCA has some limitations, its potential is deserve expecting. We believe that with the follow-up research and exploration, this method would have more widespread applications.