Refining Fitness Functions for Search-Based Automated Program Repair: A Case Study with ARJA and ARJA-e
Created by W.Langdon from
gp-bibliography.bib Revision:1.8010
- @InProceedings{Guizzo:2021:SSBSE,
-
author = "Giovani Guizzo and Aymeric Blot and James Callan and
Justyna Petke and Federica Sarro",
-
title = "Refining Fitness Functions for Search-Based Automated
Program Repair: A Case Study with {ARJA} and {ARJA-e}",
-
booktitle = "SSBSE 2021",
-
year = "2021",
-
editor = "Una-May O'Reilly and Xavier Devroey",
-
volume = "12914",
-
series = "LNCS",
-
pages = "159--165",
-
address = "Bari",
-
month = "11-12 " # oct,
-
publisher = "Springer",
-
note = "Winner Challenge Track",
-
keywords = "genetic algorithms, genetic programming, genetic
improvement, SBSE, APR, Search-based automated program
repair, empirical study, Software engineering",
-
isbn13 = "978-3-030-88105-4",
-
xxx = "https://www.dropbox.com/s/rrkogd2bqmzuslb/SSBSE2021.pdf",
-
URL = "https://discovery.ucl.ac.uk/id/eprint/10131848/",
-
DOI = "doi:10.1007/978-3-030-88106-1_12",
-
video_url = "https://www.youtube.com/watch?v=lcpYTv1TaE8",
-
code_url = "https://figshare.com/s/35ea3fd819e737ed806b",
-
size = "6 pages",
-
abstract = "Automated Program Repair (APR) strives to
automatically fix faulty software without
human-intervention. Search-based APR iteratively
generates possible patches for a given buggy software,
guided by the execution of the patched program on a
given test suite (i.e., a set of test cases).
Search-based approaches have generally only used
Boolean test case results (i.e., pass or fail), but
recently more fined-grained fitness evaluations have
been investigated with promising yet unsettled results.
Using the most recent extension of the very popular
Defects4J bug dataset, we conduct an empirical study
using ARJA and ARJA-e, two state-of-the-art
search-based APR systems using a Boolean and a
non-Boolean fitness function, respectively. We aim to
both extend previous results using new bugs from
Defects4J v2.0 and to settle whether refining the
fitness function helps fixing bugs present in large
software.
In our experiments using 151 non-deprecated and not
previously evaluated bugs from Defects4J v2.0, ARJA was
able to find patches for 6.62percent (10/151) of bugs,
whereas ARJA-e found patches for 7.24percent (12/151)
of bugs. We thus observe only small advantage to using
the refined fitness function. This contrasts with the
previous work using Defects4J v1.0.1 where ARJA was
able to find adequate patches for 24.2percent (59/244)
of the bugs and ARJA-e for 43.4percent (106/244). These
results may indicate a potential overfitting of the
tools towards the previous version of the Defects4J
dataset.",
-
notes = "https://conf.researchr.org/track/ssbse-2021/ssbse-2021-rene---replications-and-negative-results#event-overview",
- }
Genetic Programming entries for
Giovani Guizzo
Aymeric Blot
James Callan
Justyna Petke
Federica Sarro
Citations