Architecture metrics: a servey of the potential to improve softwar

Software development hinges on well-crafted architecture, a foundational phase that influences the entire system. This study investigates using architectural design metrics, which quantify various aspects of architecture. We conducted a systematic mapping study, analysing existing literature to understand how these metrics are currently employed. Our findings reveal a limited yet promising landscape. Metrics exist to assess design stability, understandability, modularity, and security. Notably, several studies explore gauging the "option-generation ability" of an architecture and its potential to generate different design choices. However, a gap exists in directly using these metrics for fault prediction. Existing research primarily focuses on the "reverse" effect, where metrics from later development stages (like code) are used to identify architectural issues. Overall, this study highlights the underutilised potential of architectural design metrics. While current research demonstrates the effectiveness of a relatively simple set of metrics for various purposes, further exploration is warranted. Future efforts should delve into data accumulation and investigate models for using these metrics for predictive purposes, ultimately enhancing software quality and development processes.