Applying Metrics to Identify and Monitor Technical Debt Items during Software Evolution

Abstract

The Technical Debt (TD) metaphor has been used in the software community as a way to manage and communicate the long-term consequences that some technical decisions may cause. Although intuitive, researches in TD do not discuss practical approaches to identify and monitor TD items, which could be applied in a transparent way during the process of software evolution. This work proposes a technique based on software metrics that automates the process of TD identification and monitoring. For that end, a subset of metrics was analyzed and related to aspects of TD items. As a form to validate the technique, it was used to analyze a TD item from a past 7 years multinational project. Evidences suggest that TD items can be related to software metrics so that software metric tools could be an important resource to automate the identification and monitoring of TD items.