Architectural smells are symptoms of bad code or design that can cause different quality problems, such as faults, technical debt, or difficulties with maintenance and evolution. Thus, it is very important to be able to understand how they evolved in the past and to predict their future evolution. In this project we propose a different machine learning models which predict architectural smells based on dependency features. The work is divided into two parts: in the first part we analyzed a single version of 60 projects at the class and package level; in the second part we analyzed 10 versions of 10 projects at the class level. All the analysis steps were implemented on the KNIME Analytics platform.

For all details please read the pdf "Prediction of architectural smells".