Deriving acceptance tests from high-level, natural language requirements that achieve full coverage is a major manual challenge at the interface between requirements engineering and testing. Conditional requirements (e.g., "If A or B then C.") imply causal relationships which - when extracted - allow to generate these acceptance tests automatically. This paper presents a tool from the CiRA (Causality In Requirements Artifacts) initiative, which automatically processes conditional natural language requirements and generates a minimal set of test case descriptions achieving full coverage. We evaluate the tool on a publicly available data set of 61 requirements from the requirements specification of the German Corona-Warn-App. The tool infers the correct test variables in 84.5% and correct variable configurations in 92.3% of all cases, which corroborates the feasibility of our approach.
Cite
@article{arxiv.2310.08234,
title = {CiRA: An Open-Source Python Package for Automated Generation of Test Case Descriptions from Natural Language Requirements},
author = {Julian Frattini and Jannik Fischbach and Andreas Bauer},
journal= {arXiv preprint arXiv:2310.08234},
year = {2023}
}
Comments
Tenth International Workshop on Artificial Intelligence and Requirements Engineering