An Empirical Comparison of Dynamic Impact Analysis Algorithms

  • Authors:

  • Alessandro Orso;Taweesup Apiwattanapong;James Law;Gregg Rothermel;Mary Jean Harrold

  • Affiliations:

  • Georgia Institute of Technology;Georgia Institute of Technology;Oregon State University;Oregon State University;Georgia Institute of Technology

  • Venue:
  • Proceedings of the 26th International Conference on Software Engineering
  • Year:
  • 2004

Quantified Score

Hi-index 0.00

Visualization

Abstract

Impact analysis 驴 determining the potential effects ofchanges on a software system 驴 plays an important rolein software engineering tasks such as maintenance, regressiontesting, and debugging. In previous work, two new dynamicimpact analysis techniques, CoverageImpact andPathImpact, were presented. These techniques performimpact analysis based on data gathered about program behaviorrelative to specific inputs, such as inputs gatheredfrom field data, operational profile data, or test-suite executions.Due to various characteristics of the algorithms theyemploy, CoverageImpactand PathImpactare expectedto differ in terms of cost and precision; however, there havebeen no studies to date examining the extent to which suchdifferences may emerge in practice. Since cost-precisiontradeoffs may play an important role in technique selectionand further research, we wished to examine these tradeoffs.We therefore designed and performed an empirical study,comparing the execution and space costs of the techniques,as well as the precisions of the impact analysis results thatthey report. This paper presents the results of this study.