Empirical Analysis on Effectiveness of NLP Methods for Predicting Code Smell

A code smell is a surface indicator of an inherent problem in the system, most often due to deviation from standard coding practices on developer’s part during development phase. Studies observe that smells made more susceptible call for modifications and corrections than did not contain smells. Restructuring at early stage saves exponentially increasing amount effort it would require address issues stemming presence these Instead using traditional features detect smells, we use user comments (given packages’ repositories) manually construct predict We three Extreme learning machine kernels over 629 packages identify eight by leveraging feature engineering aspects sampling techniques. Our findings indicate radial basis functional kernel performs best out methods with mean accuracy 98.52.