The Impact of Label Space Partitioning in Multi-label Code Smell Detection

Key Points

• Label space partitioning significantly improves detection accuracy in multi-label scenarios, highlighting its role in software quality assurance.
• The method achieved a 20% increase in accuracy compared to traditional detection techniques, marking a substantial improvement.
• Assessment using machine learning algorithms in a diverse dataset showcased its effectiveness in distinguishing code smells.
• This approach may enable better tools for developers, promoting cleaner code and improving software maintenance.