Investigating on the relationships between design smells removals and refactorings

Software systems continually evolve and this conducts to its architectural degradation due to the existence of numerous design problems. The presence of Design Smells is the main indicator of such problems, it points out the use of constructs that generally hurt system evolution. In this work, an investigation on Design Smells removals has been performed, focusing specifically on the co-occurrence of refactoring and related changes performed on a software system. An empirical study has been conducted considering the evolution history of 5 software systems. The detection of instances of multiple Design Smell types has been performed, along with all the history of the systems, along with, the detection of refactoring activities. The empirical study shows that Design Smells removals are not correlated to the presence of refactoring. The analysis provides useful indications about the percentage of activities conducted on smelly classes, including refactoring (even if these activities in few cases lead to effective smell removals).