Smells like Teen Spirit: Improving Bug Prediction Performance Using the Intensity of Code Smells

Code smells are symptoms of poor design and implementation choices. Previous studies empirically assessed the impact of smells on code quality and clearly indicate their negative impact on maintainability, including a higher bug- proneness of components affected by code smells. In this paper we capture previous findings on bug-proneness to build a specialized bug prediction model for smelly classes. Specifically, we evaluate the contribution of a measure of intensity of code smells (i.e., a measure that captures the severity of the smell), by adding it to a structural metric-based bug prediction models and comparing the results of the new model against the baseline model. Results indicate that the accuracy of a bug prediction model increases by adding the intensity of smells as predictor. Moreover, we complement this analysis by evaluating the actual gain provided by the intensity metric with respect to the other structural metrics in the model, including the ones used to compute the measure of intensity. We observe that the intensity index results much more important with respect to other metrics used for predicting the buggyness of smelly classes. Finally, we report additional analysis aimed at showing (i) to what extent false positive smell instances detected by an automatic smell detection tool impact the performance and (ii) the contribution of the intensity index in prediction models relying on process metrics or a combination of structural and process metrics.