Predicting Source Code Changes by Mining Change History
IEEE Transactions on Software Engineering
ACM SIGKDD Explorations Newsletter
An improved methodology on information distillation by mining program source code
Data & Knowledge Engineering
Hierarchical Clustering for Software Architecture Recovery
IEEE Transactions on Software Engineering
Clustering for Monitoring Software Systems Maintainability Evolution
Electronic Notes in Theoretical Computer Science (ENTCS)
A desiderata for refactoring-based software modularity improvement
Proceedings of the 3rd India software engineering conference
Leveraging design rules to improve software architecture recovery
Proceedings of the 9th international ACM Sigsoft conference on Quality of software architectures
Comparison and evaluation of source code mining tools and techniques: A qualitative approach
Intelligent Data Analysis
| Hi-index | 0.00 |
Program comprehension is an important part of software maintenance, especially when program structure is complex and documentation is unavailable or outdated. Data mining can produce structural views of source code thus facilitating legacy systems understanding.This paper presents a method for mining association rules from code aiming at capturing program structure and achieving better system understanding. A tool was implemented to assess this method. It inputs data extracted from code and derives association rules. Rules are then processed to abstract programs into groups containing interrelated entities. Entities are grouped together if their attributes participate in common rules. The abstraction is performed at the function level, in contrast to other approaches, that work at the program level.The method was evaluated using real, working programs. Programs are fed into a code analyser which produces the input needed for the mining tool. Results show that the method facilitates program comprehension by only using source code where domain knowledge andreliable documentation are not available or reliable.