Abstract

PEREZ, Boris R.. Architectural technical debt: an identification strategy. Ing. compet. [online]. 2023, vol.25, n.3, e-21413071.  Epub Sep 30, 2023. ISSN 0123-3033.  https://doi.org/10.25100/iyc.v25i3.13071.

Architectural Technical Debt is a metaphor for actions made by architects to achieve short-term goals while potentially harming the system’s long-term health. Architectural Technical Debt is difficult to detect since it is associated with a system’s long-term maintenance and evolution. In this research, we describe an architectural evolution-based method for debt identification that is backed by a supervised machine learning model and is based on information obtained from artifacts produced during architecture design. We discovered that even with a small amount of data, the machine learning model produces good results in terms of Recall and even Accuracy. The trial provides insights that allow us to conclude that this idea works well and might be utilized as a starting point to assist architects in identifying and managing Architectural Technical Debt.

Keywords : Architectural technical debt; Identification strategy; Machine learning; Software architecture.

        · abstract in Spanish     · text in English     · English ( pdf )