CEGSet: Collection of standardized cause-effect graph specifications

Krupalija, Ehlimana; Cogo, Emir; Bećirović, Šeila; Prazina, Irfan; Pozderac, Damir; Besic, Ingmar

doi:10.1109/meco58584.2023.10155063

Cited by 3 publications

(1 citation statement)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Jang and Kim later developed a software tool 27 for automatically generating a cause-effect graph from the textual representation of a system, localized to the Korean language. We recently collected CEGSet, 28 a dataset containing 65 cause-effect graph specifications, and applied machine learning methods 29 in order to successfully train a model for detecting the feasibility of cause-effect graph elements.…”

Section: Cause-effect Graphsmentioning

confidence: 99%

Usage of cause‐effect graphs for reliability analysis of an embedded IoT alarm software system

Cogo,

Karabegović

et al. 2024

Quality & Reliability Eng

Self Cite

View full text Add to dashboard Cite

Software is a part of most complex systems used for different purposes in today's world. The reliability of these systems must be verified before putting them to use, because their malfunction may lead to catastrophic failure, the loss of human lives, or huge costs. Software reliability can be verified by performing black‐box testing, without knowledge of the complex internal structure of these systems. In this work, an embedded IoT alarm system for detection of intruders and potential robbery is proposed. It is composed of three subsystems: embedded IoT, web‐server, and mobile phone subsystems. Cause‐effect graphing technique was applied for generating three types of black‐box test suites, by using the forward‐propagation and two different minimization algorithms. These test suites were used for black‐box testing the system, which was performed after injecting a maximum of 12 faults into the system in 100 iterations. The resulting software quality and reliability metrics showed that the forward‐propagation algorithm always achieves maximum fault detection rate (FDR), and this algorithm should be used when the highest demand for reliability must be met. The optimized minimization should only be used in cases where testing is very expensive, because it did not find faults in all testing iterations. The basic optimization retained high FDR, while reducing the impact of the minimized test suite and achieved a significant increase of test effect coverage values.

show abstract

Section: Cause-effect Graphsmentioning

confidence: 99%