Design pattern recovery in object-oriented software

Author(s):  
G. Antoniol ◽  
R. Fiutem ◽  
L. Cristoforetti
Author(s):  
Gary P. Moynihan ◽  
Bin Qiao ◽  
Matthew E. Elam ◽  
Joel Jones

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.


2019 ◽  
Vol 10 (3) ◽  
pp. 1-9
Author(s):  
Komal M. Birare

This article is an introduction to design patterns. Patterns are recent software engineering problem-solving discipline that emerged from the object-oriented community. The primary purpose of the pattern is communicating design insights and making patterns coherent and easier to understand. On the basis of a review of existing frameworks and the authors own experiences building visualization software, they present a series of design patterns for the domain of information visualization. The authors discuss the structure, factors use, and association of patterns bridge of data representation, graphics, and interaction.


Author(s):  
Gary P. Moynihan ◽  
Bin Qiao ◽  
Matthew E. Elam ◽  
Joel Jones

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.


2014 ◽  
Vol 971-973 ◽  
pp. 2667-2670
Author(s):  
Ke Tan Chen

Teachers file management system with advanced applications. This paper describes the functional goals of the system , the system consists of eight modules, design pattern is an object-oriented programming staff to a formal representation to solve programming problems , and the system user interface design .


Design Patterns are one of the demonstrated reusable answers for the normally experienced design issues. The identification of design pattern is significant action that underpins re-building procedure and gives insights to the designer. The uncovering of these design patterns help understand the object oriented models clearly by analyzing the relations present in the model. Many design pattern identification approaches have been proposed in past years. These methodologies work upon the behavioral, structural and semantic analysis of the software. Many algorithms were used to recognize design patterns in software. In this paper, we will be extracting an attribute relational matrix from the graph using object oriented approach. The aim of the paper is to discover all the design patterns present in the system design.


2018 ◽  
Vol 7 (2.20) ◽  
pp. 350
Author(s):  
Tapan Kant ◽  
Manjari Gupta ◽  
Anil Kumar Tripathi ◽  
Meeta Prakash

Meta-patterns are a sort of basic object-oriented constructs that have been used to design an object-oriented framework. It has been used to precisely describe possible design pattern of a framework at meta-level to manifest framework hot-spots and its corresponding adaptability. The present study is an attempt to develop a genetic algorithm approach for detecting the types and numbers of meta-patterns. For this purpose we have converted the UML class diagram of object-oriented framework and meta-patterns into directed graph and applied hybrid genetic algorithm. The obtained results from the proposed algorithm are further validated manually with the recall and precision percentage of 86.20 and 80.64 respectively. Overall the study demonstrates the utility of the uniquely proposed algorithm for the near accurate identification of meta-patterns for high reusability. This can be applied on frameworks for assessing the evolution process, documentation of hot-spots and reducing the customization effort. 


Sign in / Sign up

Export Citation Format

Share Document