Towards a strategy and tool support for test generation based on good software testing practices: classification and prioritization

• Ano de defesa: 2025
• Autor(a) principal: Villota Ibarra, Camilo
• Orientador(a): Rizzo Vincenzi, Auri Marcelo
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: eng
• Instituição de defesa: Universidade Federal de São Carlos; Câmpus São Carlos
• Programa de Pós-Graduação: Programa de Pós-Graduação em Ciência da Computação - PPGCC
• Palavras-chave em Inglês: Good software testing practices; Test cases; Test case development practices; Test case generation tools
• Área do conhecimento CNPq: CIENCIAS EXATAS E DA TERRA::CIENCIA DA COMPUTACAO::METODOLOGIA E TECNICAS DA COMPUTACAO
• Link de acesso: https://hdl.handle.net/20.500.14289/22666

This thesis addresses the challenge of defining, validating, and operationalizing good software testing practices, with a focus on improving the quality of test cases. It begins with a systematic literature review, which identifies 131 practices from 103 primary studies, refined into a set of 40 essential best practices. These practices were classified into code-oriented and no-code-oriented groups and validated through a practitioner-centered survey involving experienced software testers, who evaluated their clarity, relevance, applicability, and prioritization. Based on this foundation, the thesis introduces TAI-EvalGenTCS. This AI-based tool uses the OpenAI GPT-4 turbo model to evaluate test cases against proposed practices and generate optimized test cases. The tool was experimentally validated using 16 real-world Java projects, demonstrating its effectiveness in improving the modularity, maintainability, and compliance of the test cases with best practices. The survey results provided empirical confirmation of the proposed classification and offered valuable insights into tool support needs and industry perceptions. The contributions of this research include a structured and validated set of testing practices, a ranked list of essential test case features, empirical validation from industry practitioners, and a practical AI-powered solution that bridges the gap between test automation and test design quality.