A Change Proposal Driven Approach for Changeability Assessment Using FCA-Based Impact Analysis

Given a change proposal, how can we evaluate the changeability of the original system to absorb this change proposal before change implementation? Changes to software often have unexpected ripple effects. To avoid this and alleviate the risk of performing undesirable changes, a predictive measurement of these ripple effects should be conducted and a decision of acceptance or rejection should be made on this change proposal. In this paper, we propose an approach to evaluate a software system´s changeability with two steps. First, our approach uses formal concept analysis to perform change impact analysis ($CIA$), which estimates the ripple effects of the change proposal. Then, we propose a novel impactness metric to indicate the system´s changeability to absorb this change proposal. Case studies on three real-world programs show the effectiveness of our changeability assessment approach.