Concern Metrics for Modularity-Oriented Modernizations

Abstract

A known problem in legacy systems is the presence of crosscutting concerns in their architecture hampering and increasing the maintenance costs. A possible solution for that is the conduction of modularity-oriented modernization aiming at restructuring the legacy software system in order to improve its artifacts (source code, database, others) quality. Architecture-Driven Modernization (ADM) is focused on modernizing legacy software systems by using the concepts of model-driven architecture, software reengineering and standard metamodels. Knowledge Discovery Metamodel (KDM) is a standard from ADM that is able to represent software systems by means of KDM instances that are going to be modernized. An intrinsic part of the modernization process is to measure both KDM versions, the legacy and the modernized one. The measurement of legacy KDM instances enables the software engineer to quantify the existing problems while the measurement of the modernized one enables to verify whether the problems have been solved or not. Structured Metrics Metamodel (SMM) is a metamodel that can be used to specify metrics to be applied on KDM instances. However, even though most of the well known metrics are supported by SMM, there is no study that investigate how SMM could be used to specify concern metrics. We present how SMM can be used to specify concern metrics and a tool to support this measurement process, enabling the conduction of modularity-oriented modernizations to help ensuring the quality of the modernization process. Furthermore, we also discuss some challenges to be overcome involving the quality measurement process in ADM blueprint.

Publication
Quality of Information and Communications Technology