ABSTRACT Introduction Requirements engineering (RE) faces challenges due to the handling of increasingly complex software systems. These challenges can be addressed using generative artificial intelligence (GenAI). Given that GenAI‐based RE has not been systematically analyzed in detail, this review examines the related research, focusing on trends, methodologies, challenges, and future work directions. Methods A systematic methodology for paper selection, data extraction, and feature analysis is used to comprehensively review 238 articles published from 2019 to 2025 and available from major academic databases. Results Although generative pretrained transformer models dominate current applications (67.3% of studies), the research focus remains unevenly distributed across RE phases, with analysis (30.0%) and elicitation (22.1%) receiving the most attention and management (6.8%) remaining underexplored. Three core challenges—reproducibility (66.8%), hallucinations (63.4%), and interpretability (57.1%)—form a tightly interlinked triad affecting trust and consistency, and strong correlations ( co‐occurrence) indicate that these challenges must be addressed holistically. Industrial adoption remains nascent, with > 90% of studies corresponding to early‐stage development and only 1.3% reaching production‐level integration. Evaluation practices show maturity gaps, limited tool/dataset availability, and fragmented benchmarking approaches. Conclusions Despite the transformative potential of GenAI‐based RE, several barriers hinder its practical adoption. The strong correlations among core challenges demand specialized architectures targeting interdependencies rather than isolated solutions. The limited real‐world deployment reflects systemic bottlenecks in generalizability, data quality, and scalable evaluation methods. Successful adoption requires coordinated development across technical robustness, methodological maturity, and governance integration. A multiphase research roadmap emphasizing evaluation infrastructure strengthening, governance‐aware development, and industrial‐scale standardization is proposed.

With the rapid advancement of virtual reality (VR) technology, understanding developer discussions is essential for improving software quality and maintenance. This study is the first to systematically investigate how developer concerns across major VR platforms, namely SteamVR, Meta, and HTCVive, align with the ISO/IEC 25010 international software quality standard. We collected and analyzed 392,590 posts from 47,280 developers, using topic modeling and manual coding to map discussions to nine ISO/IEC 25010 quality characteristics. We further examined topic distributions, sentiment trends, and interaction patterns across platforms. Our findings show that developers are most challenged by interaction, compatibility, and functionality issues, emphasizing the need to enhance user experience, enable cross device integration, and maintain system stability. Discussion on performance has decreased, signaling a shift in priorities. Platform specific challenges also emerged, highlighting the need for tailored strategies for different VR ecosystems. Based on these insights, we suggest strategies that prioritize optimizing user interaction (e.g., intuitive controls, seamless navigation), strengthening cross platform compatibility (e.g., universal SDKs, shared asset pipelines), and implementing sustainable maintenance practices (e.g., clear codebases, regular updates) to foster a more robust VR software ecosystem.

Multi-View Modeling Framework for ML Systems (M3S) provides a framework to synchronize the experimental nature of machine learning and the deterministic side of traditional software engineering. However, understanding the framework's generality and limitations still requires further investigation. This paper compares the existing validation case study to a new case study of the OCT image diagnosis support system. The comparison between the two case studies shows M3S capability to handle variations in the nature of ML system analysis. However, the framework's capability to handle different ML tasks other than multi-class classification still requires further investigation.

In this study, we consider projects for developing service systems using machine learning (ML) techniques. These projects involve collaboration between various stakeholders. Several types of models representing system architectures are introduced so that stakeholders can develop a common understanding of these projects. In addition, metamodels are constructed by combining ML service systems models to provide project practitioners with a holistic view of the projects. In certain cases, these metamodels need to be extended to incorporate other business models for the business–IT alignment used in enterprises. For such situations, an enterprise architecture-based metamodel and method for managing the metamodel are proposed in this study, which provide a holistic view of business–IT alignment for ML projects. We confirm the effectiveness of the proposed metamodel and management method through real examples.

Abstract Development of machine learning (ML) systems differs from traditional approaches. The probabilistic nature of ML leads to a more experimentative development approach, which often results in a disparity between the quality of ML models with other aspects such as business, safety, and the overall system architecture. Herein the Multi-view Modeling Framework for ML Systems (M3S) is proposed as a solution to this problem. M3S provides an analysis framework that integrates different views. It is supported by an integrated metamodel to ensure the connection and consistency between different models. To facilitate the experimentative nature of ML training, M3S provides an integrated platform between the modeling environment and the ML training pipeline. M3S is validated through a case study and a controlled experiment. M3S shows promise, but future research needs to confirm its generality.

Machine learning system analysis requires different approaches for each different task and domain. Selecting a proper set of analytic models can be challenging for a specific problem. This paper discusses the extensibility of the Multi-View Modeling Framework for ML Systems approach using process mapping and extensible metamodel. We conducted a case study to evaluate the feasibility of such extensibility by extending the approach to facilitate an activity-driven analysis for an optical character recognition system. Based on the result of the case study, we found that Multi-View Modeling Framework for ML Systems is likely to be extensible.

University-level software engineering education faces the challenge of providing both fundamental concepts while delivering to their students the latest trend in tools and practices. However, software engineering programs may not be capable of solving those challenges with their own resources. In this paper, we present our experience in solving those challenges by cooperating with an industrial partner by developing a collaboration program to provide knowledge of the latest industrial software engineering practice. We discovered that the program has several other benefits besides providing knowledge of industrial software engineering practice. However, challenges and concerns still need to be solved and addressed to ensure the proper execution of the collaboration program.

Machine learning-based system requires specific attention towards their safety characteristics while considering the higher-level requirements. This study describes our approach for analyzing machine learning safety requirements top-down from higher-level business requirements, functional requirements, and risks to be mitigated. Our approach utilizes six different modeling techniques: AI Project Canvas, Machine Learning Canvas, KAOS Goal Modeling, UML Components Diagram, STAMP/STPA, and Safety Case Analysis. As a case study, we also demonstrated our approach for lane and other vehicle detection functions of self-driving cars.

Aligning business goals and strategies to software requirement is becoming more critical as corporate relies more on software for their business activities. While GQM+Strategies gives the solution to this problem, GQM+Strategies does not explicitly offer attention to relationships between various stakeholders. We propose an integration of Balanced Scorecard's perspectives into GQM+Strategies framework to solve that problem. We evaluated the possibilities by classifying goals and strategies of three existing grids totaling 73 goals and 127 strategies. We also analyzed the relationship between those perspectives in those grids. We found that current application of GQM+Strategies followed balanced scorecard's principles of perspective and concluded that it is possible to use balanced scorecard's perspectives on GQM+Strategies framework.