Distributing test cases more evenly in adaptive random testing

Abstract

Adaptive Random Testing (ART) has recently been proposed to enhance the failure-detection capability of Random Testing. In ART, test cases are not only randomly generated, but also evenly spread over the input domain. Various ART algorithms have been developed to evenly spread test cases in different ways. Previous studies have shown that some ART algorithms prefer to select test cases from the edge part of the input domain rather than from the centre part, that is, inputs do not have equal chance to be selected as test cases. Since we do not know where the failure-causing inputs are prior to testing, it is not desirable for inputs to have different chances of being selected as test cases. Therefore, in this paper, we investigate how to enhance some * A preliminary version of this paper was presented at the 18th Australian Software Engineering Conference (ASWEC 2007) (Chen et al., 2007b

Does Adaptive Random Testing Deliver a Higher Confidence than Random Testing?

“…In this paper, we work on two particular ART algorithms, namely fixed-size-candidate-set ART (FSCS-ART) [7] and an enhancement of FSCS-ART, namely FSCS-ART with partitioning by edge and centre (ECP-FSCS-ART) [5]. These ART algorithms and their failure-detection capabilities are described as follows.…”

“…They further pointed out that the edge preference is a cause of the deterioration of the failure-detection capability of FSCS-ART for high dimensional cases. In order to offset the edge preference of the original FSCS-ART, a new algorithm ECP-FSCS-ART [5] has been proposed. ECP-FSCS-ART works as follows.…”

“…c 2 will finally be selected as the next test case. The detailed algorithm of ECP-FSCS-ART can be found in [5]. Some simulation studies in [5] have shown that ECP-FSCS-ART significantly improves the failure-detection capability of FSCS-ART, especially for the cases of high dimension.…”

Combinatorial testing, random testing, and adaptive random testing for detecting interaction triggered failures

“…RT has the following advantages: (1) low overhead, easy to implement and automate; (2) lightweight, requiring little information from the software specification and source code; (3) generates test cases that are less likely created by human testers, testing software without any bias so as to detect critical software failures that are not considered by human testers nor covered by deterministic approaches [23]; and (4) quantitative estimates of the software's operational reliability can be inferred from random testing [24].…”

Diagnostic Path-Oriented Test Data Generation by Hyper Cuboids

“…Test data generation, is done by the different methods such as genetic algorithm [4,5,6], randomly [7,8,9], etc. The test data generators are divided into categories of random generator, data specification generator, and path-oriented generator [10,11].…”