An Adaptive Approach for Fault Localization using R-CNN

Software accuracy and dependability become very important issues now-a-days. It is more difficult to identify software program errors due to the growing size and complexity of programs. Traditional fault localization methods fall short of finding all the defects in a huge real-world program. The use...

Full description

Saved in:
Bibliographic Details
Published in2022 International Conference on Advancements in Smart, Secure and Intelligent Computing (ASSIC) pp. 1 - 6
Main Authors Ghosh, Debolina, Singh, Jay Prakash, Singh, Jagannath
Format Conference Proceeding
LanguageEnglish
Published IEEE 19.11.2022
Subjects
Benchmark testing
Complexity theory
Data models
Debugging
Fault Localization
Location awareness
Machine learning
R-CNN
Software
Software Testing
Training
Online AccessGet full text

Cover

More Information
Summary:Software accuracy and dependability become very important issues now-a-days. It is more difficult to identify software program errors due to the growing size and complexity of programs. Traditional fault localization methods fall short of finding all the defects in a huge real-world program. The use of machine learning in that situation is effective. In this research, we present an R-CNN model-based fault localization technique. The coverage matrix is used as the model's input during training, and it uses test data to determine each program statement's suspiciousness score. Each statement is given a rank based on its score. Three benchmark programs have been looked at to evaluate the proposed work's reliability. The analysis of the results demonstrates that the suggested method can effectively identify false statements.
DOI:10.1109/ASSIC55218.2022.10088417