A technique to evaluate software evolution based on architecture metric

Abstract

Software evolution is always happening during its lifetime. In a software evolution process, the change in software structure often leads to software quality degradation, makes it difficult to maintain or transfer to other platform. In this paper, we propose a technique to evaluate software evolution based on architecture metric. We split the whole architecture evolution process into a series of atomic evolution operation steps, analyze the impact of each atomic change operation with examples, and then find out the general evolution trend. Our purpose is to analyze how architecture changes influence the relevant software quality attributes, which helps to maintain good software quality and keep software healthily evolving.