Ciências e Tecnologia | Comunicações em congressos, conferências, seminários/Communications in congresses, conferences, seminars
URI permanente para esta coleção:
Navegar
Entradas recentes
- Static recording apparatus for soundscape analysis in MacauPublication . Estadieu, Gerald; Cheung, Y. Y.; Pestana, Pedro Duarte; Barbosa, ÁlvaroThe concept of Soundscape was initially proposed to study the relationship between humans and their sonic environment. It has gathered momentum from academia to environmentalists and policymakers throughout the years. The study and characterisation of Soundscapes can be complex as it tries to take a holistic and qualitative approach rather than simply quantifying sound pressure levels. This paper introduces a comprehensive Soundscape study process in an ongoing research project in Macao (China), a small territory (32.9 km2) and one of the most densely populated areas in the world. The paper seeks to show a first version of a technical solution to systematically capture the local soundscape, analyse it, classify it, and ultimately deliver a dataset library and the intangible qualities of the environmental sound. This implementation, including technical documentation, code, and sound library with strong labelling, is presented under an open-source license to encourage future collaborative research. Finally, the paper offers suggestions on further developing the apparatus to reach a systematic and near real-time soundscape analysis with the development of a machine learning system.
- Enhancing competency development and organizational effectiveness through advanced technologies: a position paperPublication . Dias, J. T.; Santos, Arnaldo; Martins, P.; São Mamede, HenriqueIn recent years, companies have faced increasing pressure from glob-alization, requiring them to adapt not only to survive but also to thrive in a highly competitive environment. This adaptation has been facilitated by the ef-ficient integration of technology, achieved through digital processes and collab-oration tools. Digital transformation has emerged as a critical element for main-taining competitiveness as economies become increasingly digital. To succeed in this ever-evolving environment, companies must balance leveraging existing strengths with seeking new organizational agility. Integrating advanced tech-nologies like Artificial Intelligence (AI) and Web Technologies, into education and professional training is a strategic response to the challenges posed by the current digital landscape. AI, with its adaptability and automation capabilities, offers benefits such as increased efficiency, personalized learning, and stream-lined administrative processes. Continuous evaluation of teaching and learning, along with data extraction and predictive analysis, enhances e-learning quality and informs organizational decisions. This research aims to investigate how ad-vanced technologies can predict and adapt organizational training needs to im-prove competency development and overall effectiveness. The research adopts a Design Science Research (DSR) methodology, focusing on the development and implementation of an AI-based framework for personalized training rec-ommendations. Expected outcomes include integrating AI-driven predictive models with existing Human Resources Management Systems to identify and address training needs, fostering employee skill development, organizational agility, and competitiveness in a rapidly changing market. Additionally, ad-dressing this issue promotes a more inclusive and empowering work environ-ment, enabling employees to thrive in an increasingly digital world.
- Pylung: a supporting tool for comparative study of ViT and CNN-based models used for lung nodules classificationPublication . Marques, Felipe; Pestana, Pedro Duarte; Filipe, VitorLung cancer is a significant global health concern, and accu- rate classification of lung nodules plays a crucial role in its early detec- tion and treatment. This paper evaluates and compares the performance of Vision Transformer (ViT) and Convolutional Neural Network (CNN) models for lung nodule classification using the Pylung tool proposed in this work. The study aims to address the lack of research on ViT in lung nodule classification and proposes ViT as an alternative to CNN. The Lung Image Database Consortium and Image Database Resource Ini- tiative (LIDC-IDRI) dataset is utilized for training and evaluation. The Pylung tool is employed for dataset preprocessing and comparison of models. Three models, ViT, VGG16, and ResNet50, are analyzed, and their hyperparameters are optimized using Optuna. The results show that ViT achieves the highest accuracy (99.06%) in nodule classifica- tion compared to VGG16 (98.71%) and ResNet50 (98.46%). The study contributes by introducing ViT as a model for lung nodule classification, presenting the Pylung tool for model comparison, and suggesting further investigations to improve the accuracy.
- Professional training for effective adoption of Generative AI in the corporate world: bridging the gapPublication . Guedes, Fernando; Rocio, Vitor; Martins, PauloThis position paper emphasizes the critical role of professional training in facilitating the effective adoption of Generative AI (GenAI) in the corporate world. GenAI, with its ability to create new content from existing data, holds immense potential for transforming business processes, enhancing decision-making, and driving innovation. However, the adoption of GenAI faces significant challenges, including a shortage of skilled professionals, high implementation costs, data privacy concerns, and the complexity of integrating these technologies into existing systems. To address these challenges, this paper highlights the importance of comprehensive education and training programs tailored to equip employees with the necessary skills and knowledge. Such programs should focus on developing technical competencies and understanding the operational implications of GenAI. By analyzing current literature and case studies, this paper identifies key strategies for effective training and outlines best practices for integrating GenAI into corporate environments. The findings underscore the need for a strategic approach to training that aligns with the evolving demands of AI-driven innovation. This includes continuous learning and development initiatives, the promotion of a culture of innovation, and the implementation of responsible AI practices. By investing in professional training, organizations can bridge the skills gap, mitigate risks, and fully leverage the transformative potential of GenAI technologies, ultimately gaining a competitive edge in the market. Through this comprehensive exploration, the paper advocates for the integration of robust training frameworks that support the sustainable adoption of GenAI, ensuring that businesses are well-prepared to navigate the complexities and opportunities of the digital age.
- Enhancing digital libraries through NLP and recommender systems: current trends and future prospects with large language modelsPublication . Cardoso, Heike da Silva; Rocio, VitorIn an era characterized by rapid proliferation of scientific publications and overwhelming volumes of digital content, researchers, students, and faculty members face significant challenges in identifying literature relevant to their academic pursuits. This saturation of information has heightened the need for advanced Recommender Systems within university libraries, tailored specifically for navigating and discovering scientific literature. This paper proposes leveraging insights from librarians’ direct interactions with users to adapt existing Recommender Systems, augmented with NLP and LLMs, to better serve the specific needs of academic researchers. It should streamline the research process by delivering precise, relevant, and personalized literature recommendations, centered on a curated database of bibliographic information.
- Exploring the effects of crustacean fishing on cold-water coral and gorgonian ecosystems: conservation strategies for Carbon sequestrationPublication . Seixas, Sónia; Parrinha, Joaquim; Gomes, PedroCrustacean fishing has significant impacts on cold-water coral and gorgonian communities. These coldwater corals include Vulnerable Marine Ecosystems (VMEs) such as Dendrophyllia ramea and D. cornigera. Our observations indicate that fishing gear used to capture crustaceans like lobster, brown crab, and European crab along the western coast of Portugal is frequently abandoned or anchored in rocky areas. When these traps are brought aboard, they often retrieve cold-water corals and gorgonians attached to them. Sometimes, only small branches are collected, while entire structures may come to the surface on other occasions. Fishermen tend to break the corals into smaller pieces to avoid damaging their gear. These organisms play a crucial role in carbon fixation, yet they face significant destruction. The only effective way to mitigate this destruction is through educational initiatives aimed at fishermen, encouraging the return of salvaged corals and gorgonians to the sea whenever possible. Furthermore, any corals that cannot be returned should be taken ashore for proper restoration and rehabilitation, followed by subsequent reintroduction into their natural habitat. Implementing such measures is essential for minimising the impact on carbon sequestration dynamics and safeguarding the region's overall biodiversity.
- Sistema de Classificação de Sinalética Gestual em competições de karatéPublication . Violante, Sónia Correia; Filipe, Vítor; Morais, A. Jorge; Rocha, Álvaro; Peñalvo, Francisco; Gonçalves, Ramiro; Garcia Holgado, Alicia; Moreira, FernandoEm contexto de Kumite (combate de Karate) propõe-se investigar um modelo de classificação da sinalética gestual do árbitro para atribuição de pontos, com recurso a Visão Computacional e técnicas de Aprendizagem Profunda. Foram realizadas três abordagens, todas tendo como base o recurso a modelos de Redes Neuronais Convolucionais (Convolutional Neural Network – CNN): Classificação de imagens com recurso a uma CNN; Deteção da pose humana com o modelo MoveNet; e a deteção e classificação de gestos com o modelo YOLOv5, via RoboFlow. A última abordagem obteve melhores resultados, com 100% de precisão para todas as classes, pelo que se testou a sua aplicação para a deteção e classificação dos gestos em vídeo.
- Potential for trophic transfer of microplastics in the smallspotted catshark’s food web: insights from Portuguese coastal ecosystemsPublication . Pica, Rodrigo; Fernandes, João; Seixas, Sónia; Martins, Marta; Greife, Anna; Müller, Carolin; Heel, LenaMicroplastics (MPs) are increasingly prevalent pollutants in marine environments, raising concerns about their potential to transfer across trophic levels. However, the extent and significance of such transfer remain uncertain. This study aimed to (i) characterise MPs in the digestive tracts (GITs) of three ecologically connected species — the small-spotted catshark (Scyliorhinus canicula), pouting (Trisopterus luscus), and Henslow’s swimming crab (Polybius henslowii); and (ii) assess the potential for trophic transfer, as S. canicula is a known predator of the other two species. Thirty individuals from each species were collected by bottom-trawling off Figueira da Foz, Portugal. GITs were digested using 10% potassium hydroxide (KOH), filtered, and analysed by Fourier-transform infrared (FTIR) spectroscopy. A total of 88 MPs were identified, primarily in S. canicula (0.89 ± 1.19 MP.ind-1) and P. henslowii (0.54 ± 0.74 MP.ind-1). No MPs were found in T. luscus, possibly due to its juvenile stage or short retention time. Frequencies of occurrence were 57% for S. canicula and 43% for P. henslowii, representing the first record of MPs in the latter. The predominant MP types were fragments (51%) and fibres (47%), mainly green (32%), black (20%), and transparent (16%), with polyvinyl acetate (32%) and polyethene terephthalate (27%) identified as the main polymers. Significant differences in MP colour and polymer composition between species suggest distinct exposure routes. Although the presence of MPs in a known prey species of S. canicula suggests potential trophic transfer, further research is needed to confirm bioaccumulation. These findings provide a valuable baseline for understanding plastic pollution in coastal ecosystems and highlight the need for ongoing investigation into MP dynamics within marine food webs.
- Large language model for querying databases in PortuguesePublication . Figueiredo, Lourenço; Pinheiro, Paulo; Cavique, Luís; Marques, NunoThis study introduces a system that helps non-expert users find information easily without knowing database languages or asking technicians for help. A specific domain is explored, focusing on a subscription-based sports facility, which serves as an open-source version of a real case study. Utilizing the star schema, the available data in the database is structured to provide accessibility through Portuguese Natural Language queries. Using a Large Language Model (LLM), SQL queries are generated based on the question and the provided star schema. We created a dataset with 115 highly challenging questions drawn from real-world usage scenarios to validate the correctness of the system. Challenges found during testing, like attribute value interpretation, out-of-scope questions, and temporal interval adequacy issues, highlight the insufficiency of the star schema alone in providing the needed context for generating accurate SQL queries by the LLM. Addressing these challenges through enhanced contextual information shows significant improvement in query correctness, with validation results increasing from 57.76% to 88.79%. This study shows the potential and limitations of LLMs in generating SQL queries from Portuguese Natural Language queries.
- Data science maturity model: from raw data to pearl’s causality hierarchyPublication . Cavique, Luís; Pinheiro, Paulo; Mendes, Armando B.Data maturity models are an important and current topic since they allow organizations to plan their medium and long-term goals. However, most maturity models do not follow what is done in digital technologies regarding experimentation. Data Science appears in the literature related to Business Intelligence (BI) and Business Analytics (BA). This work presents a new data science maturity model that combines previous ones with the emerging Business Experimentation (BE) and causality concepts. In this work, each level is identified with a specific function. For each level, the techniques are introduced and associated with meaningful wh-questions.We demonstrate the maturity model by presenting two case studies.