http://repositorio.ufc.br/handle/riufc/19757 2026-06-10T05:59:04Z http://repositorio.ufc.br/handle/riufc/86626 Título: Graph attention network para reconhecimento de emoções a partir de eletroencefalogramas Autor(es): Nascimento, Vanessa Carvalho do Abstract: The automatic recognition of emotional states is relevant in several areas, such as mental health, brain-computer interfaces, and affective monitoring systems. Electroencephalography (EEG) is a non-invasive technique for recording brain electrical activity that stands out in this context for directly reflecting the neural states associated with emotions, being harder to mask than facial expressions or voice. However, characteristics such as the non-stationary nature of the signals, low signal-to-noise ratio, and inter-subject variability make this task challenging. In this scenario, Graph Neural Networks (GNNs) stand out for directly modeling the relationships between EEG channels. Thus, this work proposes a model based on Graph Attention Network (GAT), a GNN variant that adaptively weights the contribution of each neighboring channel during information aggregation, for the classification of three emotional states (positive, neutral, and negative) present in the SJTU Emotion EEG Dataset (SEED). Each EEG sample is represented as a graph whose nodes correspond to EEG channels and whose edges connect pairs of channels whose cosine similarity between their respective feature vectors exceeds the 70th percentile of the sample similarity distribution. Among the 62 channels and 5 frequency bands conventionally used in the literature, only 4 channels (FT7, FT8, T7, and T8) and 2 bands (delta and theta) were selected, resulting in a compact configuration with potential application in wearable devices with a limited number of electrodes. Experiments were conducted under the Leave-One-Subject-Out (LOSO) protocol, which evaluates generalization across subjects. The model achieves a mean accuracy of 95.38% over 10 independent runs, outperforming many comparable works that use the full channel and band configuration, indicating that graph-based modeling captures relevant discriminative patterns even with a reduced configuration. Tipo: Dissertação 2026-01-01T00:00:00Z http://repositorio.ufc.br/handle/riufc/86625 Título: NaVISOL: desenvolvimento de um chatbot com sistema de recomendação para o mapa digital da cidade de Sobral Autor(es): Ramos, Alex de Sousa Abstract: The growing public demand for fast, clear, and efficient access to public services and information, especially in urban contexts increasingly dependent on digital technologies, makes the development of solutions that facilitate citizen interaction with complex databases, such as Geographic Information Systems (GIS), essential. The availability of large volumes of information on interactive mapping platforms, although promising, still faces usability limitations, often resulting in navigation difficulties, cognitive overload, and underutilization of available resources. In this context, this work proposes the development of the Navegador Virtual Inteligente de Sobral (NaVISOL), a conversational agent based on Artificial Intelligence (AI) and Natural Language Processing (NLP), implemented using the Rasa framework and integrated into the Sobral em Mapas platform, with the objective of enabling a more accessible interface for querying and exploring geospatial data. The system incorporates a hybrid recommendation engine that combines conversational context, implicit user navigation feedback, and Singular Value Decomposition (SVD) techniques for similarity analysis among map layers, mitigating data sparsity issues and promoting proactive discovery of relevant content. To evaluate the effectiveness of the solution, quantitative analyses were conducted based on the usage incidence metric of the recommendations, as well as a qualitative assessment of user experience through questionnaires addressing usability aspects, response comprehension, and perceived relevance of the provided suggestions. The results indicate that the adoption of the conversational interface associated with the recommendation system contributes to a more intuitive interaction with GIS, increases access to recommended layers, and fosters the discovery of new geographic content, demonstrating the potential of NaVISOL as a tool to democratize access to urban information and strengthen citizen-centered digital government initiatives. Tipo: Dissertação 2026-01-01T00:00:00Z http://repositorio.ufc.br/handle/riufc/85471 Título: Uma validação das leis de Lehman: uma investigação quantitativa no desenvolvimento de software moderno Autor(es): Sousa, Augusto Braz de Abstract: This study investigates the applicability and relevance of Lehman’s Laws in the current context of software development, using a quantitative approach to analyze three modern systems: Dropwizard, K-9 Mail, and LanguageTool. Through the Goal Question Metric (GQM) framework, the research empirically validates Lehman’s Laws, which describe fundamental principles of software evolution, such as continuous change, increasing complexity, and conservation of organizational stability. The study is divided into two main stages. The first stage involves analyzing the evolution metrics of the three selected software systems, verifying their compliance with Lehman’s Laws. The second stage utilizes the lessons learned to develop practical guidelines that can be applied to software development and maintenance. The results confirm that Lehman’s Laws are applicable and relevant to modern software systems, providing a theoretical foundation for guiding development and maintenance practices. The analyzed metrics demonstrate growth and complexity patterns that corroborate Lehman’s proposed laws. Additionally, the study offers a summary of the lessons learned and practical guidelines for developers and maintenance teams, aiming to improve adaptability, reduce complexity, and ensure the continuous quality of systems. These guidelines are based on empirical evidence collected and offer recommendations for the efficient evolution of software. The main contributions of this work include the empirical validation of Lehman’s Laws, the application of the GQM framework as a methodological basis, and the creation of practical guidelines based on the lessons learned. Tipo: Dissertação 2024-01-01T00:00:00Z http://repositorio.ufc.br/handle/riufc/85469 Título: Desacoplamento de controle longitudinal e lateral em veículos autônomos usando controle preditivo baseado em modelo: validação usando o simulador CARLA Autor(es): Rodrigues, João Pedro da Silva Abstract: This dissertation presents the comparison and validation of two trajectory control strategies for autonomous vehicles within a hierarchical planning and control architecture. In the first approach, referred to as coupled control, a Model Predictive Controller (MPC) simultaneously performs both lateral and longitudinal control actions, while in the second approach, referred to as decoupled control, the predictive controller is responsible for lateral control and a proportional– integral (PI) controller governs the vehicle’s longitudinal dynamics. In both cases, the predictive control is based on the dynamic bicycle model. The lateral controller ensures that the vehicle follows the desired trajectory, while the longitudinal controller maintains the vehicle at a predefined constant speed by generating acceleration based on the longitudinal reference velocity. The reference trajectories and velocity are generated by a trajectory planner that employs a receding-horizon control strategy based on mixed-integer quadratic programming. Validation is carried out through co-simulation in MATLAB and CARLA environments, where the controllers are implemented in MATLAB and the vehicle model is simulated in CARLA, an open-source high-fidelity simulator for autonomous driving research. The simulation experiments consider two distinct scenarios. In the first scenario, the maneuver of a vehicle traveling along a one-way road in the presence of a static obstacle is evaluated. In the second scenario, the maneuver is evaluated in the presence of a dynamic obstacle moving at a constant longitudinal speed. The results show that the coupled strategy achieves better lateral trajectory tracking, especially at higher speeds, with a root mean square error (RMSE) of 0.084 m compared to 0.089 m for the decoupled method. However, it is less accurate in longitudinal speed control, with an RMSE of 1.32 m compared to 1.21 m. The computational performance analysis indicates that both approaches meet real-time requirements and are computationally feasible, with the decoupled method exhibiting a slightly lower solution time (2.6 ms versus 2.7 ms). Tipo: Dissertação 2025-01-01T00:00:00Z