http://repositorio.ufc.br/handle/riufc/533 2026-06-11T09:06:57Z 2026-06-11T09:06:57Z Carneiro, Ana Cesarina de Oliveira http://repositorio.ufc.br/handle/riufc/86489 2026-05-26T18:48:39Z 2026-01-01T00:00:00Z

Título: Realidade virtual com 0DoF no suporte ao desenvolvimento de crianças com transtorno do espectro autista Autor(es): Carneiro, Ana Cesarina de Oliveira Abstract: This study investigates the application of Virtual Reality (VR) with a 0DoF architecture as a tool to support the development of social, cognitive, and emotional skills in children with Autism Spectrum Disorder (ASD). VR enables the creation of controlled, predictable, and adaptable virtual environments, allowing children with ASD to practice social and emotional interactions in safe and structured contexts. The methodological choice of the 0DoF architecture is grounded in neurocognitive evidence related to sensory alterations present in ASD, aiming to reduce perceptual overload and increase experimental control. The research was conducted between May 2024 and February 2026 and was structured into three complementary stages: (i) a Systematic Literature Review aimed at mapping scientific evidence regarding the use of VR in the context of ASD; (ii) an exploratory study involving 130 participants, including parents, educators, and therapists, conducted through questionnaires and semi-structured interviews, with data collection carried out between November 3 and November 28, 2024, seeking to understand perceptions, expectations, and challenges related to the adoption of this technology; and (iii) the development and experimental application of a virtual environment called Island of Emotions, designed with a 0DoF architecture, prioritizing environmental predictability and reducing sensorimotor variability. The results were analyzed through descriptive quantitative analysis and qualitative thematic analysis. The findings indicate that VR, when structured in a controlled and adaptive manner, promotes engagement, contributes to emotional regulation, and stimulates the acquisition of social skills, while also reinforcing its potential as a complementary tool in therapeutic and educational contexts. Challenges related to technological accessibility, professional training, and the personalization of interventions were also identified, highlighting the importance of aligning theoretical foundations, technological architecture, and the sensory needs of the target population. Tipo: Dissertação

2026-01-01T00:00:00Z Cavalcante, João Charllys Brauwol Araújo http://repositorio.ufc.br/handle/riufc/86444 2026-05-25T15:13:54Z 2026-01-01T00:00:00Z

Título: Uma Plataforma para integração de dados de obras públicas Autor(es): Cavalcante, João Charllys Brauwol Araújo Abstract: The growth in the volume of public data on government works, available on different platforms, highlights challenges in the communicability and reliability of this information. The lack of a unified system can hinder oversight and transparency, thus compromising the efficient management of allocated resources. Given this scenario, it becomes essential to improve the integration and organization of this information to strengthen social control over public works information and decision-making by managers. The main objective of this work is to provide users with a solution aimed at transparency of data on federal public works, as well as to develop a website, as a proof of concept, dedicated to improving knowledge production and increasing transparency in this field. The results include a systematic literature review, which includes several factors that can hinder the data integration process for public works, such as heterogeneity, inconsistency, and data duplication. Scripts were also developed for extraction, standardization, integration, error monitoring, and a web interface as a proof of concept. The results highlight the proposed approach to integrating and structuring data on public works, reducing information fragmentation and improving access to information. Furthermore, a systematic review revealed the need for standardization and automation in the processing of this data. The implementation of the stated platform has the capacity to execute operations in this initial phase, being a modulebased architecture and cloud computing that facilitates the integration process. Tipo: Dissertação

2026-01-01T00:00:00Z Umbelino, Wellington de Sousa http://repositorio.ufc.br/handle/riufc/86376 2026-05-19T17:11:44Z 2026-01-01T00:00:00Z

Título: Planejamento temporal para gerenciamento de dispositivos de internet das coisas em um campus universitário Autor(es): Umbelino, Wellington de Sousa Abstract: Energy costs in public buildings in Brazil constitute a critical problem, especially in light of the consumption reduction targets established by the federal government. Several measures have been adopted to mitigate this scenario, such as awareness campaigns aimed at building occupants. However, there are still opportunities for improvement, particularly with the adoption of technologies such as the Internet of Things (IoT), which enable basic levels of automation. For a more effective approach, it is necessary to consider not only device automation, but also factors such as environment occupancy, user comfort, and variations in energy prices. Different Artificial Intelligence approaches, such as model predictive control, machine learning, and automated planning, have been explored to address this problem. In this context, this work proposes the application of the Temporal Planning technique for the management of IoT devices, through the integration between a communication system of an IoT network and a planning system capable of performing automated reasoning. As a result, an architecture is proposed that enables the control of electrical equipment in a university campus at the Federal University of Ceará. Tipo: Dissertação

2026-01-01T00:00:00Z Martins, Francisco Leonardo Batista http://repositorio.ufc.br/handle/riufc/86151 2026-05-07T13:51:59Z 2025-01-01T00:00:00Z

Título: Avaliando o desempenho de grandes modelos de linguagem na realização de provas de dedução natural em lógica proposicional e lógica de predicados Autor(es): Martins, Francisco Leonardo Batista Abstract: The growing use of conversational agents has sparked increasing interest among researchers, educators, and educational institutions worldwide. Moreover, with the recent popularization of Large Language Models (LLMs), numerous studies have been conducted to explore the use of these tools to support the teaching and learning process. These systems are capable of understanding and processing vast amounts of data, which allows them, for example, to provide individualized support to students in solving exercises. However, it is essential to consider that such systems generally employ reasoning processes based on statistical methods, such as machine learning algorithms, which may produce incorrect answers in tasks requiring logical reasoning. This study aims to evaluate the ability of Large Language Models (LLMs) to solve Natural Deduction exercises in Propositional Logic and Predicate Logic. To this end, experiments were carried out using the GPT 4.1-mini, GPT 4o, and GPT 3.5turbo models to solve natural deduction exercises. These experiments were conducted a priori, without training, and subsequently, further experiments were performed on the trained GPT 4.1-mini model to assess potential performance improvements. The results indicate a significant improvement after training, although the model still exhibits a considerable number of errors. Therefore, if employed for this task, its use is recommended with due caution. Tipo: Dissertação

2025-01-01T00:00:00Z