Before interacting with real users, developers must be proficient in human–computer interaction (HCI) so as not to exhaust user patience and availability. For that, substantial training and practice are required, but it is costly to create a variety of high-quality HCI training materials. In this context, chat generative pretrained transformer (ChatGPT) and other chatbots based on large language models (LLMs) offer an opportunity to generate training materials of acceptable quality without foregoing specific human characteristics present in real-world scenarios. Personas is a user-centered design method that encompasses fictitious but believable user archetypes to help designers understand and empathize with their target audience during product design. We conducted an exploratory study on the Personas technique, addressing the validity and believability of interviews designed by HCI trainers and answered by ChatGPT-simulated users, which can be used as training material for persona creation. Specifically, we employed ChatGPT to respond to interviews designed by user experience (UX) experts. Two groups, HCI professors and professionals, then evaluated the validity of the generated materials considering quality, usefulness, UX, and ethics. The results show that both groups rated the interviews as believable and helpful for Personas training. However, some concerns about response repetition and low response variability suggested the need for further research on improved prompt design in order to generate more diverse and well-developed responses. The findings of this study provide insight into how HCI trainers can use ChatGPT to help their students master persona creation skills before working with real users in real-world scenarios for the first time.

Context : Games are a well-established scenario to test AI and Multi-Agent Systems (MAS) proposals due to their popularity and defiance. However, there is no big picture of the application of this technology to games, the evolution of the kind of problem tackled, or the game scenarios in which agents have been experimented. Objective : To perform a systematic mapping to characterise the state of the art in the field of MAS applied to virtual games and to identify trends, strengths, and gaps for further research. Method: A Systematic Mapping Study has been conducted to find primary studies in the field. A search was performed on title, abstracts, and keywords, whilst classification, data extraction, and further analysis were performed according to specific criteria focused on MAS papers with experimentation and evidence in a game scenario. Results : 78 studies published between 1998 and 2021 were found. Studies have been classified according to the MAS problem faced and the agent reasoning strategy. We detect that Machine Learning is the most common AI technique for MAS in games, considering both reinforcement learning and evolutionary techniques. MAS are used in a variety of gaming genres, especially in Real-Time Strategy (RTS), Sports and Simulation. Conclusions : RTS and Sports games are well-suited for concrete MAS problems such as multi-agent planning and task allocation. Expanding evidence and experimentation on other aspects related to scalability and usability issues is discussed. Those MAS problems and experiments that remain slightly modelled on games or are not thoroughly studied yet have been also identified.

Context: At present, sensor-based systems are widely used to solve distributed problems in changing environments where sensors are controlled by intelligent agents. On Multi-Agent Systems, agents perceive their environment through such sensors, acting upon that environment through actuators in a continuous cycle. These problems have not always been addressed from an ad-hoc perspective, designed specifically for the circumstances of the problem at hand. Instead, they have been modelled under a common mathematical framework as distributed constrained optimisation problems (DCOP). Objective: The question to answer is how sensor-based scenarios have been modelled as DCOPs in changing environments known as Dynamic DCOP and what their trends, gaps, and progression are. Method: A systematic mapping study of Dynamic DCOPs has been conducted, considering the scattered literature and the lack of consensus in the terminology. Results: Given the high complexity of distributed constraint-based problems, priority is given to obtaining sub-optimal but fast responses with a low communication cost. Other trending aspects are the scalability and guaranteeing the solution over time. Conclusion: Despite some lacks in the analysis and experimentation in real-world scenarios, a large set that is applicable to changing sensor-based scenarios is evidenced, along with proposals that allow the integration of off-the-shell constraint-based algorithms.

Frailty condition is an intermediate state in the ageing trajectory, preceding the onset of disability, and is becoming both challenging and priority to guarantee the sustainability of healthcare systems. To address it, we designed a software platform for supporting CGA and monitoring the activities performed at patients' dwelling. We have implemented an alert engine to inform the care professionals of intrinsic capacity decline, and a decision support system for referral recommendations. Besides, the platform enables them to provide tailored interventions to their patients.

The integration of Large Language Models (LLMs) with Monte Carlo Tree Search (MCTS) offers a novel approach to tackle decision-making challenges in turn-based strategy games, a domain where traditional methods face scalability and adaptability limitations. The objective is to explore the feasibility of LLMs in improving MCTS strategic reasoning by processing dynamic and probabilistic game states. Although prior research has demonstrated the strengths of MCTS in tactical optimization and the potential of LLMs in natural language reasoning, the synergy between these approaches remains underexplored. To explore this gap, this study faces the team formation problem in the Pokémon VGC AI framework through “PokeHit”, a hybrid AI agent that incorporates LLM-driven reasoning within the MCTS pipeline. By testing open-source models as Llama and Mistral, this work evaluates their ability to guide macro-strategy decisions and compares their performance against baseline-heuristic algorithms. Results reveal that LLMs can provide strategies dynamically but suffer from limitations, including hallucinations and inconsistent improvements over traditional methods. However, this research supports the feasibility and challenges of hybrid AI approaches in fine-tuning and policy design to improve their efficiency. The findings aim to contribute to the field of intelligent game agents through the applicability of LLMs in decision-making domains.