Development Of Interaction Test Data Generation Strategy With Input-Output Mapping Supports

Uniform strength t-way testing (where t represents interaction strength) forms the basis of interaction testing. However, t is rarely uniform in real world as not all interaction faults are solely constituted by these fixed t-interactions. Consequently, a general solution has been introduced: input-...

Full description

Saved in:
Bibliographic Details
Main Author: Ong, Hui Yeh
Format: Thesis
Language:English
Published: 2012
Subjects:
Online Access:http://eprints.usm.my/46327/1/ONG%20HUI%20YEH_HJ.pdf
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Uniform strength t-way testing (where t represents interaction strength) forms the basis of interaction testing. However, t is rarely uniform in real world as not all interaction faults are solely constituted by these fixed t-interactions. Consequently, a general solution has been introduced: input-output based relationship interaction testing. Although useful, most existing strategy implementations are lacking in terms of the automated input-output mapping support (to translate the symbolic outputs back into actual data form) and test suite generation flexibility. In order to address these aforementioned issues, a non-deterministic input-output based relationship interaction testing strategy, AURA, has been developed. AURA strategy also integrated with post-processing automated input-output mapping support and flexible iteration control capability to support test suite generation flexibility. Experimental results indicated that AURA strategy is generating competitive test suite size against existing strategies (Density, ParaOrder, Union, TVG, PICT, AETG, ACA, GA-N, IPO-N, IPO, Jenny, SA and ACS). Specifically, this strategy is capable to generate the test suite size as optimized as other strategies for certain inputs. Lastly, the post-processing automated input-output mapping support and flexible iteration control capability are evaluated with experiments.