Theoretical Analysis of GP-Evolved 08/05/2012 Risk Evaluation Formulas for Spectrum Based Fault Localisation (2013)
BibTeX
@MISC{Xie13theoreticalanalysis,
author = {Xiaoyuan Xie and Fei-ching Kuo and Tsong Yueh Chen and Shin Yoo and Mark Harman},
title = {Theoretical Analysis of GP-Evolved 08/05/2012 Risk Evaluation Formulas for Spectrum Based Fault Localisation},
year = {2013}
}
Share
 |
 |
 |
 |
||
OpenURL
Abstract
Spectra-Based Fault Localisation (SBFL) aims to assist de- bugging by applying risk evaluation formulæ (sometimes called suspiciousness metrics) to program spectra and ranking statements according to the predicted risk. Designing a risk evaluation formula is often an intuitive process done by human software engineer. This paper presents a Genetic Programming approach for evolving risk assessment formulæ. The empirical evaluation Risk evaluation formulas convert program spectrum data from test executions into suspiciousness score, according to which statements are ranked to aid debugging. activities. GP-evolved Designing equations can consistently such formulas outperform remained largely many a of manual the task human-designed until Genetic Programming formulæ, has such beenas recently Tarantula, Ochiai, Jaccard, applied: resulting Ample, formulas and Wong1/2, showed promising up to 5.9 performance times. More in empirical importantly, evaluation. they can We in- perform equally vestigate as well the as GP-evolved Op2, which formulas was theoretically recently proved and prove to that be GP optimal has produced against fourIf-Then-Else-2 maximal (ITE2) structure, formulas that or even had not outperform been knownit before. against More other interestingly, program some structures. of the newly found maximal formulas show characteristics that may seem inconsistent with human intuition. This is the first SBSE result with provable human competitiveness. 1 The program spectra data used in the paper, as well as the complete empirical results, are available from:
Keyphrases
risk evaluation formula fault localisation theoretical analysis program spectrum data empirical evaluation risk evaluation formula gp-evolved designing equation intuitive process risk assessment formul gp-evolved op2 genetic programming formul program spectrum formula outperform human intuition performance time fourif-then-else-2 maximal spectra-based fault localisation test execution suspiciousness metric risk evaluation formul complete empirical result maximal formula provable human competitiveness first sbse result human software engineer suspiciousness score genetic programming approach
