A Multi-Objective Genetic Algorithm to Test Data Generation

Evolutionary testing has successfully applied search based optimization algorithms to the test data generation problem. The existing works use different techniques and fitness functions. However, the used functions consider only one objective, which is, in general, related to the coverage of a testi...

Full description

Saved in:
Bibliographic Details
Main Authors: Pinto, G H L, Vergilio, S R
Format: Conference Proceeding
Language:English
Subjects:
Context
Genetics
Java
Memory management
Optimization
Software
Testing
Citations: Items that cite this one
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Evolutionary testing has successfully applied search based optimization algorithms to the test data generation problem. The existing works use different techniques and fitness functions. However, the used functions consider only one objective, which is, in general, related to the coverage of a testing criterion. But, in practice, there are many factors that can influence the generation of test data, such as memory consumption, execution time, revealed faults, and etc. Considering this fact, this work explores a multiobjective optimization approach for test data generation. A framework that implements a multi-objective genetic algorithm is described. Two different representations for the population are used, which allows the test of procedural and object-oriented code. Combinations of three objectives are experimentally evaluated: coverage of structural test criteria, ability to reveal faults, and execution time.
ISSN:1082-3409
2375-0197
DOI:10.1109/ICTAI.2010.26