Use Case Mining Based On Formal Concept Analysis

Use case is very important to understanding, maintenance, transplant -ation and reengineering of legacy system. At present, research on mining use cases focus on examining and analyzing the code of object-oriented system. However, many legacy systems were developed before the appearing of object-oriented design method, their codes were usually hard to obtain, and most of those codes which can be gotten also cannot be analyzed to precisely obtain the needs of business functions of system.This thesis focuses on mining use cases from application system's actual movement behavior, by checking out the use of the legacy system which has rich user interface and frequent system-user interaction, using the formal concept analysis technique. The thesis treats records of the system-user interaction as basic sources of information for extracting use cases of legacy system. User completes system's service through the application system, so the mined use cases reflect the needs of business functions of system.After introducing the mining method of use cases at home and abroad with the present development, formal concept analysis and algorithm of constructing concept lattice, this thesis proposes a method of mining use cases based on formal concept analysis, discusses the construction of concept lattices based on system-user interaction, implements improved algorithm of constructing concept lattice, and mining system use cases through horizontal decomposition and analysis of concept lattice.Finally, this thesis uses "postgraduate work management systems" as a experiment platform, explains the process in detail from the construction and decomposition of concept lattice to the final excavation of a system's use cases, and uses relevant data to prove the feasibility of this method of study.