Software Validation by Means of Statistical Testing: Retrospect and Future Direction

Thévenod-Fosse, Pascale

doi:10.1007/978-3-7091-9123-1_2

Cited by 13 publications

(8 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The alternative bound on N of N > (2/p)(2 + In(I/p)) for QN~(I-p), proposed in [27,28], has two drawbacks when compared with (2): the parameters QNand p are not set independently, and for a given p, it greatly overestimates N [29].…”

Section: Theoretical Workmentioning

confidence: 99%

“…The three steps identified in the strategy apply to each phase of the validation process, from unit level to system level. The main complementary features of deterministic and random test inputs with respect to verification and dependability evaluation are summarized in Figure I (see [29] for a complete discussion). Figure 1 corresponding to the percentages of revealed faults.…”

Section: Validation Strategymentioning

confidence: 99%

“…The authors believe that the second ordering is more cost effective provided that 'an appropriate input distribution is used to generate random data. The three steps of the proposed strategy are as follows [29]: step 1: statistical testing for fault removal. Test data are randomly selected based on an input probability distribution that aims at minimizing the number N of executions required to achieve a target test quality; step 2: deterministic testing for fault removal.…”

Section: Principle Of the Strategymentioning

confidence: 99%

See 2 more Smart Citations

An investigation of statistical software testing

Thévenod-Fosse

Waeselynck

1991

Software Testing Verif & Rel

View full text Add to dashboard Cite

Software validation embodies two notions: fault removal and fault forecasting. Statistical testing involves exercising a piece of software by supplying it with input values that are randomly selected according to a defined probability distribution on its input domain. It can be used as a practical tool for revealing faults in a fault removal phase, and for assessing software dependability in a fault forecasting phase. In both of these, its efficiency is linked to the adequacy of the input probability distribution with respect to the test experiment goal. In this paper a mixed validation strategy combining deterministic and random test data is defined, and the theoretical and experimental work performed to support the strategy is reported. The quoted results relate to the unit testing of four real programs from the nuclear field. They confirm the high fault revealing power of statistical structural testing. Two main directions for further investigation of statistical testing are indicated by the reported work. These are described and solutions to the associated problems are outlined.

show abstract

Section: Theoretical Workmentioning

confidence: 99%

Section: Validation Strategymentioning

confidence: 99%

Section: Principle Of the Strategymentioning

confidence: 99%

See 1 more Smart Citation

An investigation of statistical software testing

Thévenod-Fosse

Waeselynck

1991

Software Testing Verif & Rel

View full text Add to dashboard Cite

show abstract

“…A notion of test quality for statistical testing methods has been defined first in [30]. We slightly reformulate it for our context.…”

Section: Coverage Criteria and Randomnessmentioning

confidence: 99%

“…In [30], q C,N (A) is called the test quality of the method with respect to C. The test quality q C,N (A) can be easily stated if q C,1 (A) is known. Indeed, one gets q C,N (A) = 1−(1−q C,1 (A)) N , since when drawing N tests, the probability of reaching an element is one minus the probability of not reaching it N times.…”

Section: Coverage Criteria and Randomnessmentioning

confidence: 99%

Coverage-biased random exploration of large models and application to testing

Denise

Gaudel

Gouraud

et al. 2011

Int J Softw Tools Technol Transfer

View full text Add to dashboard Cite

Abstract. This article presents several related methods for drawing traces. First, it is shown how to draw traces uniformly at random in large models composed of several components. Then a method for drawing traces according to a given coverage criterion is presented, together with a notion of randomised coverage satisfaction. These methods rely on combinatorial algorithms, based on a representation of the model by an automaton or by a product of several automata, synchronised or not. We report several experimental results on random generation of traces in large transition systems, and on statistical testing of C programs.

show abstract

A random testing approach using pushdown automata

Dreyfus

Héam

Kouchnarenko

et al. 2014

Software Testing Verif & Rel

View full text Add to dashboard Cite

International audienceSince finite automata are in general strong abstractions of systems, many test cases which are automata traces generated uniformly at ran-dom, may be un-concretizable. This paper proposes a method extending the abovementioned testing approach to pushdown systems providing finer abstractions. Using combinatorial techniques guarantees the uniformity of generated traces. In addition, to improve the quality of the test suites, the combination of coverage criteria with random testing is investigated. The method is illustrated within both structural and model-based testing contexts

show abstract

Software Validation by Means of Statistical Testing: Retrospect and Future Direction

Cited by 13 publications

References 17 publications

An investigation of statistical software testing

An investigation of statistical software testing

Coverage-biased random exploration of large models and application to testing

A random testing approach using pushdown automata

Contact Info

Product

Resources

About