tradeoff analysis

Existing architecture evaluation methods (e.g., ATAM) focus on concrete systems and struggle with the generality and variability challenges arising from multi-level abstractions—software architectures, reference architectures, and architecture frameworks. This paper proposes ATRAF, a scenario-driven unified assessment framework, introducing the first cross-level trade-off analysis paradigm. It comprises three complementary methods: ATRAM (for software architecture), RATRAM (for reference architecture), and AFTRAM (for architecture framework), collectively supporting holistic quality attribute trade-offs (e.g., modifiability, performance, security) and risk identification. Leveraging iterative, spiral-style scenario modeling, sensitivity analysis, and feedback-driven optimization, ATRAF refines and extends ATAM. Evaluated on an RTS case study, ATRAF demonstrates robust incremental assessment capability, significantly improving early architectural decision quality and cross-level consistency assurance.

This work addresses the longstanding reliance on manual, time-consuming, and subjectivity-prone approaches in software architecture quality assessment, particularly in scenarios involving trade-offs among multiple quality attributes. To overcome these limitations, the study introduces generative large language models (LLMs) into this domain for the first time, leveraging Microsoft Copilot in conjunction with the Architecture Tradeoff Analysis Method (ATAM) and quality attribute scenario techniques. The proposed approach enables automated identification of architectural risks, analysis of sensitivity points, and generation of trade-off recommendations. Experimental results demonstrate that, in most cases, the method achieves higher accuracy and efficiency compared to human-led reviews, substantially reducing evaluation costs while improving consistency across assessments.

This study addresses the labor-intensive, time-consuming, and subjective nature of quality attribute scenario selection and trade-off analysis in Architecture Tradeoff Analysis Method (ATAM) evaluations. It pioneers the integration of a commercial large language model (LLM)—Microsoft Copilot—into pedagogical ATAM practice to support risk identification, sensitivity point analysis, and cross-attribute trade-off reasoning. Methodologically, a structured prompt engineering framework is developed to formalize key ATAM process steps into LLM-executable tasks. Empirical evaluation across multiple case studies demonstrates that the LLM-generated outputs—covering risks, sensitivity points, and trade-off assessments—achieve higher accuracy than initial human assessments and substantially reduce evaluation turnaround time. The primary contributions are: (1) empirical validation of commercial LLMs’ efficacy and feasibility in architectural quality scenario analysis; and (2) provision of a reusable technical pathway and evidence-based foundation for advancing both ATAM automation and architecture education.

This work addresses the challenges of low quality and poor transparency in build-or-buy decisions within enterprise software development, which often stem from reliance on unstructured experiential knowledge. To overcome these limitations—particularly in cold-start scenarios lacking historical data—the authors propose a structured approach that integrates a decision-factor ontology, rule-based reasoning, and reference-class matching. This method enables transparent, auditable evaluation of alternatives and represents the first application of combined ontology modeling and rule reasoning to build-or-buy decision-making. By revealing critical decision thresholds and supporting traceability, the approach enhances the rationality, transparency, and auditability of choices. Its practical efficacy is demonstrated through a lightweight tool validated in a financial industry case study, showing significant improvements in decision quality.

This paper addresses the misalignment between industrial practice (ATRAF framework) and academic conventions (IMRaD format) in software architecture evaluation research. We propose a systematic alignment method that maps ATRAF’s four-phase spiral model onto the IMRaD structure, enabling bidirectional traceability between evaluation processes and scholarly writing norms. Our approach introduces a unified, iterative workflow spanning all abstraction levels—integrating scenario-driven view modeling, multi-attribute sensitivity analysis (e.g., performance, modifiability, security), and quantitative risk assessment. This enhances rigor, transparency, and reproducibility of architecture evaluation studies. Empirical validation across multiple academic case studies demonstrates the method’s expressive power and practical utility, effectively bridging the gap between industrial evaluation practices and scholarly reporting standards.

This work proposes a lightweight framework to address the high cost and poor contextual retention inherent in traditional engineering decision capture methods. By modeling decision alternatives as slices of system models and embedding them directly into model-based systems engineering (MBSE) workflows, the approach explicitly links decision knowledge with requirements, behavioral elements, and architectural components. This integration significantly reduces the overhead of decision documentation while enhancing the preservation of contextual information. The feasibility of the framework is demonstrated through a case study on aircraft architecture simplification, which confirms its effectiveness in improving decision reusability and integration efficiency within MBSE environments.

This study addresses the challenge of quantifying the complexity and cost induced by external requirement changes when detailed knowledge of a system’s internal logic is unavailable. To this end, the authors propose a black-box assessment method based on a directed graph of component coupling. By analyzing component interfaces and integrating multi-view modeling—graphical, algebraic, and tabular—the approach uniquely links interface characteristics to cost factors, enabling computable bounded estimates of change-induced complexity and associated costs. The method was validated through a large-scale integration case in a retail banking platform, demonstrating its effectiveness and providing architects and operations teams with actionable, quantitative insights for system design and maintenance.

This work addresses the challenge of security assessment in cyber-physical systems caused by missing or outdated architectural documentation. It proposes ASTRAL, a novel approach that uniquely integrates multimodal large language models with architectural modeling to automatically reconstruct system architectures from fragmented data. By leveraging prompt chaining, few-shot learning, and architectural reasoning, ASTRAL enables the identification of attack surfaces and supports quantitative analysis of risk propagation pathways. Evaluated on multiple real-world systems, the method demonstrates strong efficacy and has received endorsement from 14 cybersecurity experts, significantly enhancing the reliability and decision-support capability of architecture-driven security assessments.