Automatic Human-Like Detection of Code Smells

Many code smell detection techniques and tools have been proposed, mainly aiming to eliminate design flaws improve software quality. Most of them are based on heuristics which rely a set metrics corresponding threshold values. Those suffer from subjectivity issues, discordant results among the tools, reliability thresholds. To mitigate these problems, we used machine learning automate developers’ perception in smells detection. Different other existing trained our models with an extensive dataset more than 3000 professional reviews 518 open source projects. We conclude by empirical evaluation performance approach against PMD, widely metric-based tool for Java. The experimental show that outperforms PMD classifier all evaluations.