Identification of reusable components within an object-oriented software system using algebraic graph theory

Xanthos, Spiros

doi:10.1145/1028664.1028784

Cited by 1 publication

(1 citation statement)

References 3 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Shi [9] extracting GoF design patterns to gain better perspective on the design level of software systems. Xanthos [10] uses graph partitioning methods for finding reusable software subsystems in a software.…”

Section: Related Workmentioning

confidence: 99%

Software design metric based analysis of dependency patterns

Sarica

Ovatman

2013

2013 Second International Conference on Informatics &Amp; Applications (ICIA)

View full text Add to dashboard Cite

Graph based analysis is a widely known and applied technique in understanding large and complicated software designs. For instance when analyzing class diagrams, the number of classes and relations can reach numbers where straightforward inspection becomes infeasible to detect meaningful class(es) in the design. Compacting the diagram by defining patterns that form clusters of nodes can become handy in solving such a problem. Dependency patterns are one of the many approaches that address this issue. However, when it comes to object oriented designs it is necessary to assign further meaning to dependency graph based structures. In this study, we provide metric based correlation analysis to investigate the meaning of key patterns that can be found in the dependency graphs extracted from UML class diagrams. We extracted dependency patterns of many modern object oriented software and apply correlation analysis on the metric values of the classes that reside in those patterns. Our results show that some of the metrics are significantly more correlated than the others for certain pattern types. Our findings can be useful in detecting meaningful classes (or class groups) whose properties are identified by the relation between those metrics.

show abstract

Section: Related Workmentioning

confidence: 99%