22 May 2023 in Software and Systems Modeling
Most experts agree that large language models (LLMs), such as those used by Copilot and ChatGPT, are expected to revolutionize the way in which software is developed. Many papers are currently devoted to analyzing the potential advantages and limitations of these generative AI models for writing code. However, the analysis of the current state of LLMs with respect to software modeling has received little attention. In this paper, we investigate the current capabilities of ChatGPT to perform modeling tasks and to assist modelers, while also trying to identify its main shortcomings. Our findings show that, in contrast to code generation, the performance of the current version of ChatGPT for software modeling is limited, with various syntactic and semantic deficiencies, lack of consistency in responses and scalability issues. We also outline our views on how we perceive the role that LLMs can play in the software modeling discipline in the short term, and how the modeling community can help to improve the current capabilities of ChatGPT and the coming LLMs for software modeling.
17 August 2022 in Software and Systems Modeling
8 April 2024 in IEEE Software
The use of generative AI chatbots in formative assessment can effectively gauge learning progress, increase the academic performance of students compared to a traditional methodology, and raise student awareness about the tradeoffs of employing generative AI in their work.
1 December 2018 in Science of Computer Programming
23 January 2014 in Journal of the Brazilian Computer Society
Self-adaptive systems are software-intensive systems endowed with the ability to respond to a variety of changes that may occur in their environment, goals, or the system itself by adapting their structure and behaviour at run-time in an autonomous way. Controllers are complex components incorporated in self-adaptive systems, which are crucial to their function since they are in charge of adapting the target system by executing actions through effectors, based on information monitored by probes. However, although controllers are becoming critical in many application domains, so far very little has been done to assess their robustness. In this paper, we propose an approach for evaluating the robustness of controllers for self-adaptive software systems, aiming to identify faults in their design. Our proposal considers the stateful nature of the controller and identifies a set of robustness tests, which includes the provision of mutated inputs to the interfaces between the controller and the target system (i.e. probes). The feasibility of the approach is evaluated on Rainbow, a framework for architecture-based self-adaptation, and in the context of the Znn.com case study.