The article discusses mutation testing, a technique in software engineering that introduces syntactic changes to program code to create high-utility test suites. It evaluates the effectiveness of this approach by assessing how well test suites detect these changes. The study highlights the significance of generating high-utility mutations to reduce computational costs during testing and explores the impacts of various prompts and language models in generating these mutations. It also investigates the causes of non-compilable mutations, underscoring the need for improved methodologies in mutation generation for better software reliability.
Mutation testing provides an effective strategy to create high-utility test suites by introducing syntactic modifications to code, thereby identifying weaknesses in test suites.
High-utility mutations are critical for effective mutation testing, as they minimize the computational costs associated with compiling and executing each mutation.
The study explores the impacts of different prompts and generative models in mutation generation, analyzing how they affect behavior similarity and usability.
Identifying the root causes and types of non-compilable mutations is essential for improving mutation generation techniques in software engineering.
Collection
[
|
...
]