Current challenges in automatic software repair

The abundance of defects in existing software systems is unsustainable. Addressing them is a dominant cost of software maintenance, which in turn dominates the life cycle cost of a system. Recent research has made significant progress on the problem of automatic program repair, using techniques such...

Full description

Saved in:
Bibliographic Details
Published in:Software quality journal 2013-09, Vol.21 (3), p.421-443
Main Authors: Le Goues, Claire, Forrest, Stephanie, Weimer, Westley
Format: Article
Language:English
Subjects:
Algorithms
Automation
Compilers
Computer programs
Computer Science
Data Structures and Information Theory
Defects
Evolutionary algorithms
Focusing
Instrumentation
Interpreters
Operating Systems
Programming Languages
Repair
Run time (computers)
Software
Software engineering
Software Engineering/Programming and Operating Systems
Studies
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:The abundance of defects in existing software systems is unsustainable. Addressing them is a dominant cost of software maintenance, which in turn dominates the life cycle cost of a system. Recent research has made significant progress on the problem of automatic program repair, using techniques such as evolutionary computation, instrumentation and run-time monitoring, and sound synthesis with respect to a specification. This article serves three purposes. First, we review current work on evolutionary computation approaches, focusing on GenProg, which uses genetic programming to evolve a patch to a particular bug. We summarize algorithmic improvements and recent experimental results. Second, we review related work in the rapidly growing subfield of automatic program repair. Finally, we outline important open research challenges that we believe should guide future research in the area.
ISSN:0963-9314
1573-1367
DOI:10.1007/s11219-013-9208-0