Student research abstract: A methodology to predict the instable classes

Class stability is intrinsically characterized by the evolution of a number of dependencies and change propagation factors used to promote the ripple effect. In this regard, historical information regarding change propagation factors can aid to identify the classes prone to ripple effect (that is instable classes). In this paper, we propose a methodology to exploit the versions history of change propagation factors in order to predict the instable classes. Initially, we have implemented the proposed methodology with version history of three open source projects MongoDB Java Driver, Google Guava and Apache MyFaces and obtained promising results as compared to existing stability assessors. Subsequently, the experimental results indicate that proposed methodology can be used to identify the classes prone to ripple effect and can aid developers to reduce the efforts needed to maintain and evolve the system. Copyright is held by the owner/author(s).