http://hdl.handle.net/10174/157 2026-04-06T05:54:47Z http://hdl.handle.net/10174/41415 Title: Are Small Language Models Enough for Biomedical QA Tasks? Authors: Lamar-Leon, Javier; Nogueira, Vitor; Quaresma, Paulo Abstract: This paper presents a specialized fine-tuning approach for the Mistral-7B Large Language Model (LLM) tailored for biomedical applications. We employ Low-Rank Adaptation (LoRA), a parameter-efficient fine-tuning method, to adapt the model to the intricacies of biomedical language and domain-specific knowledge. By integrating LoRA, we aim to preserve the general language understanding capabilities of Mistral-7B while enhancing its performance on biomedical tasks. The fine-tuning process involves training the model on the PubMedQA dataset. Our experiments demonstrate that the fine-tuned Mistral-7B model achieves notable accuracy, 60%. This performance is particularly significant given the relatively modest size of the Mistral-7B model compared to other approaches that often require larger models to achieve comparable results. The results highlight the effectiveness of LoRA in fine-tuning large language models for domain-specific applications, particularly in the biomedical field, where precise and contextually accurate language understanding is crucial. This work contributes to the advancement of AI in healthcare by providing a robust and efficient method for adapting LLMs to biomedical applications, demonstrating that high precision can be achieved with a smaller model size. 2025-08-17T23:00:00Z http://hdl.handle.net/10174/41322 Title: IoTCRAFT: Toward a Declarative Approach for IoT System Specification and Deployment Authors: Cardoso, João; Salgueiro, Pedro; Gonçalves, Teresa Abstract: Developing and deploying IoT systems remains a complex task, often involving the manual integration of heterogeneous compo- nents for implementing data pipelines, data storage, visualization, and machine learning workflows. This paper proposes IoTCRAFT, a declar- ative platform designed to define high-level specifications and automate the deployment of Internet of Things (IoT) systems. The platform is intended to support complete data processing pipelines, including data ingestion, validation, normalization, storage, and visualization, as well as the training and inference of machine learning models. By using a high-level specification to describe system behavior, the platform’s code generation and orchestration engine can automatically provision the re- quired infrastructure using container-based technologies. This approach aims to reduce the need for manual configuration and simplifies the de- ployment of IoT systems. 2025-08-31T23:00:00Z http://hdl.handle.net/10174/41318 Title: A Model to Manage Security Risks in IoT Systems Authors: Lento, Luiz; Patinho, Pedro; Abreu, Salvador Abstract: The IoT is an information and communication technology, which provides greater speed and flexibility in activities performed and decision-making. It is a reality that, through network connectivity and computing capacity, allows interconnected devices to generate, exchange and process data with minimal human intervention, thus offering a wide range of functionalities to the computing universe. However, this universe of options provided by IoT technology brings with it a universe of concerns regarding information security. A key issue with IoT environments is ensuring security across all services and devices. The diversity of threats, together with the lack of concern of most of its administrators and device designers, make the IoT network environment vulnerable. Therefore, this article aims to present a new strategy for improving security in IoT systems. The strategy is to manage risks dynamically, enabling their monitoring and control in real time. In this way, risks will be analyzed and evaluated dynamically, enabling security mechanisms to be applied before major damage is caused to IoT systems. 2025-01-01T00:00:00Z http://hdl.handle.net/10174/41314 Title: Um Modelo Declarativo para Gestão de Riscos em IoT Authors: Lento, Luiz; Patinho, Pedro; Abreu, Salvador Abstract: Um grande problema em ambientes IoT ´e garantir a seguranc¸a em todos os servic¸os e dispositivos. A diversidade de ameac¸as, em conjunto com a falta de preocupac¸ ˜ao da maioria de seus administradores e projetistas dos dispositivos, tornou o ambiente de rede IoT vulner´avel. Este artigo apresenta o RTRMM, um modelo de gerenciamento de riscos de seguranc¸a baseado em l´ogica para ambientes IoT, que prevˆe os riscos e visa gerenci´a-los em tempo real, tornando o ambiente IoT mais con´avel. Faz uso de probabilidade, l´ogica difusa e programac¸ ˜ao em l´ogica para implementar as suas funcionalidades. 2023-01-01T00:00:00Z