Generative AI to Generate Test Data Generators
Benoît Baudry, Khashayar Etemadi, Sen Fang, Yogya Gamage, Yi Liu, Yuxin Liu, Martin Monperrus, Javier Ron, André Silva, Deepika Tiwari
Abstract
High quality data is essential for designing effective software test suites. We propose three original methods for using large language models to generate representative test data, which fit to the domain of the program under test and are culturally adequate.
Topics & Concepts
Generative grammarTest (biology)Computer scienceArtificial intelligenceTest dataNatural language processingProgramming languagePaleontologyBiologySoftware Testing and Debugging TechniquesSoftware System Performance and ReliabilityData Mining Algorithms and Applications