Development of an Intelligent Information Systems for Object-Oriented Software Design
Gary P. Moynihan (The University of Alabama, USA), Bin Qiao (The University of Alabama, USA), Matthew E. Elam (The University of Alabama, USA) and Joel Jones (The University of Alabama, USA)
Copyright: © 2008
The purpose of this research was to apply an artificial intelligence approach to improve the efficiency of design pattern selection used in the development of object-oriented software. Design patterns provide a potential solution to the limitations occurring with traditional software design approaches. Current methods of design pattern selection tend to be intuitive, and based on the experience of the individual software engineer. This expertise is very specialized and frequently unavailable to many software development organizations. A prototype expert system was developed in order to automate this process of selecting suitable patterns to be applied to the design problem under consideration. It guides the designer through the pattern selection process through inquiry regarding the nature of the design problem. The prototype system also provides the capabilities to browse patterns, view the relationship between patterns, and generate code based on the pattern selected. The routine application of such a system is viewed as a means to improve the productivity of software development by increasing the use of accepted design patterns.