Use this URL to cite or link to this record in EThOS:
Title: Test data generation : two evolutionary approaches to mutation testing
Author: May, Peter S.
ISNI:       0000 0001 3621 9563
Awarding Body: University of Kent
Current Institution: University of Kent
Date of Award: 2007
Availability of Full Text:
Access from EThOS:
Access from Institution:
Despite nearly 30 years of research and being widely held as a powerful unit testing technique, mutation testing still suffers from a number of problems. For the most part, its hindrance lies with the large number of mutated program executions that must occur, although there are also a number of other challenges such as automation, test case generation, and identifying semantically equivalent programs. Over the years many techniques have tried to help overcome these problems; this thesis provides another, tailored specifically to the area of automatic test case generation using a biological metaphor. Nature has been regarded as a plentiful supply of ideas and metaphors for our own engineering needs. In computing terms, this practice has spawned many new algorithms primarily designed at optimisation and adaptation, with one of the most infamous being Genetic Algorithms (GA). Recently however, a new paradigm has emerged that offers promising results compared to GAs - Artificial Immune Systems (AIS). As their name suggests, these algorithms look towards the immune system to provide inspiration for solving complex and adaptive problems, often with favourable results. Genetic Algorithms have previously been applied to mutation testing in the area of test data generation, with reasonable success. This thesis compares such an approach to an immune system inspired algorithm, indicating that the latter is capable of generating higher mutation scores in lower execution times. In addition, an analysis of each algorithm's parameter space is performed, highlighting to practitioners useful settings for each parameter in respect to the program being tested, as well as possible algorithm refinements.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
Keywords: QA 76 Software, computer programming