A Study on Different Spectra in Fault Localization

Nícolas Hamparsomian, Marcos Lordello Chaim

2025

Abstract

We present an experimental study to assess the impact of different spectra in fault localization. We evaluated one machine learning-based technique (Deep Neural Networks—DNN) and two Spectrum-based fault localization (Ochiai and Tarantula). These techniques were applied on 319 faulty versions of industry-like programs with real bugs using control (statements) and data (definition use associations—DUA) flow coverage as spectra. The results suggest that DNN does not benefit from data flow spectra and any spectrum will generate similar results using either Ochiai or Tarantula. Among the techniques and spectra assessed, Ochiai using control flow seems to be the best choice for fault localization.

Paper Citation

in Harvard Style

Hamparsomian N. and Chaim M. (2025). A Study on Different Spectra in Fault Localization. In Proceedings of the 20th International Conference on Evaluation of Novel Approaches to Software Engineering - Volume 1: ENASE; ISBN 978-989-758-742-9, SciTePress, pages 492-499. DOI: 10.5220/0013278500003928

in Bibtex Style

@conference{enase25,
author={Nícolas Hamparsomian and Marcos Chaim},
title={A Study on Different Spectra in Fault Localization},
booktitle={Proceedings of the 20th International Conference on Evaluation of Novel Approaches to Software Engineering - Volume 1: ENASE},
year={2025},
pages={492-499},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0013278500003928},
isbn={978-989-758-742-9},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 20th International Conference on Evaluation of Novel Approaches to Software Engineering - Volume 1: ENASE
TI - A Study on Different Spectra in Fault Localization
SN - 978-989-758-742-9
AU - Hamparsomian N.
AU - Chaim M.
PY - 2025
SP - 492
EP - 499
DO - 10.5220/0013278500003928
PB - SciTePress