Streamlining code smells: Using collective intelligence and visualization

• Autor(a) principal: Reis, José Vicente Pereira dos
• Data de Publicação: 2022
• Idioma: eng
• Título da fonte: Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos)
• Texto Completo: http://hdl.handle.net/10071/26465
• topic: Crowdsmelling; Code smells; Crowdsourcing; Code smells detection; Machine learning; Software quality; Cheiros de código; Deteção de cheiros de código; Aprendizagem automática; Qualidade do software

Context. Code smells are seen as major source of technical debt and, as such, should be detected and removed. Code smells have long been catalogued with corresponding mitigating solutions called refactoring operations. However, while the latter are supported in current IDEs (e.g., Eclipse), code smells detection scaffolding has still many limitations. Researchers argue that the subjectiveness of the code smells detection process is a major hindrance to mitigate the problem of smells-infected code. Objective. This thesis presents a new approach to code smells detection that we have called CrowdSmelling and the results of a validation experiment for this approach. The latter is based on supervised machine learning techniques, where the wisdom of the crowd (of software developers) is used to collectively calibrate code smells detection algorithms, thereby lessening the subjectivity issue. Method. In the context of three consecutive years of a Software Engineering course, a total “crowd” of around a hundred teams, with an average of three members each, classified the presence of 3 code smells (Long Method, God Class, and Feature Envy) in Java source code. These classifications were the basis of the oracles used for training six machine learning algorithms. Over one hundred models were generated and evaluated to determine which machine learning algorithms had the best performance in detecting each of the aforementioned code smells. Results. Good performances were obtained for God Class detection (ROC=0.896 for Naive Bayes) and Long Method detection (ROC=0.870 for AdaBoostM1), but much lower for Feature Envy (ROC=0.570 for Random Forrest). Conclusions. Obtained results suggest that Crowdsmelling is a feasible approach for the detection of code smells, but further validation experiments are required to cover more code smells and to increase external validity