http://repositorio.jesuita.org.br/handle/UNISINOS/10015 Ciência da Computação Thu, 09 Apr 2026 08:55:38 GMT 2026-04-09T08:55:38Z http://repositorio.jesuita.org.br/handle/UNISINOS/11970 DASTData: um modelo baseado em Fog e Cloud Computing para rastreabilidade e armazenamento distribuído de dados IoT em Smart Cities Ferreira, Daniel Lopes In this work, a solution for the distributed storage of data in Smart Cities is pre sented. Through an Edge-Fog-Cloud architecture that partitions the data through the Sharding technique, a hierarchical model is proposed that manipulates IoT data generated by Smart Cities. The problem is related to the approaches used to promote an integrated environment. Related works tend to use cloud-focused approaches that generate high latency rates, and those that use Fog Computing only use the layer as middleware, not exploring greater possibilities for use. In this context, this work presents the DASTData model that aims to enable lower latency rates, more data security, fault tolerance, high availability and concurrent queries to promote a better experience in data management and availability in smart cities. In addition, our contribution to the literature, unlike related works, is related to the proposition of an architecture focused on enabling the traceability of users who have a mobile behavior in the city, providing the ability to analyze patterns and occurrences through the consolidation of data from one or more indi viduals. In the results obtained through the tests carried out in this work, it is observed that in queries a decentralized model such as DASTData is up to 73% more efficient than a centralized model. Fri, 24 Jun 2022 00:00:00 GMT http://repositorio.jesuita.org.br/handle/UNISINOS/11970 2022-06-24T00:00:00Z http://repositorio.jesuita.org.br/handle/UNISINOS/11969 Auxiliando o diagnóstico clínico de glaucoma a partir da segmentação e classificação automatizada de imagens de fundo de olho Ceschini, Lucas Mayer Fri, 10 Dec 2021 00:00:00 GMT http://repositorio.jesuita.org.br/handle/UNISINOS/11969 2021-12-10T00:00:00Z http://repositorio.jesuita.org.br/handle/UNISINOS/11968 Estimativa do índice de evapotranspiração com base em técnicas de aprendizado de máquina Kuhn, Fabiane In line with climate change and water crises, the population growth has drawn atten tion to agriculture. The sector is responsible for the use of 70% of the world’s water and a waste of approximately 50% of this total in irrigation processes. Several technological methods are being developed to minimize this impact and collaborate with the UN Sustainable Development Goals, within them, sustainable agriculture. Aiming at optimizing the use of water resources, it is necessary to analyze evapotranspiration, as it is the most active variable in the hydrological cycle and the main component in the water balance of agricultural ecosystems. Based on the hypothesis that Machine Learning analysis can be applied to determine evapotranspiration and aid in decision making in irrigation, allowing assertive estimates and without dependence on a wide variety of data, this article presents the application of Decision Tree Regressor techniques. , Random Forest, Artificial Neural Network and XGBoost for that purpose. Using a dataset from the National Institute of Meteorology (INMET), the models were trained based on widely validated equations for Evapotranspiration calculations. After the testing routine, it was possi ble to obtain satisfactory results, with MAE less than 0.0015, demonstrating the effectiveness of computational techniques for estimating evapotranspiration. Tue, 07 Dec 2021 00:00:00 GMT http://repositorio.jesuita.org.br/handle/UNISINOS/11968 2021-12-07T00:00:00Z http://repositorio.jesuita.org.br/handle/UNISINOS/11165 Análise de desempenho de linguagens de programação e APIs quânticas Viegas, Jorge Matheus Gomes Wed, 30 Jun 2021 00:00:00 GMT http://repositorio.jesuita.org.br/handle/UNISINOS/11165 2021-06-30T00:00:00Z http://repositorio.jesuita.org.br/handle/UNISINOS/11164 Uma proposta de modelo combinando thick data e big data para potencializar os resultados das análises de dados Sipp, Eduardo Big data is a revolution in terms of data storage. Such data allow for several analyzes, possibilities for generating valuable insights and to support decision-making, however, the vast majority of these analyzes still have a high failure rate or low profitability when observing the amount invested versus the results obtained. Considering this scenario, the present work has as the main objective to analyze how thick data combined with big data can be used to make the analyzes more assertive. In summary, thick data intends to understand human behavior and how the human relationship with a particular product or service evolves over time. Understanding such issues allow more accurate extraction of data, the execution of better analyzes and, consequently, tends to generate better results. To conduct the research, the Design Science Research (DSR) research method was used. The paper presents the proof of concept of the model developed as well as two evaluations to observe the viability of the model, one theoretical using two cases from the literature and another practical based on Twitter data and Federal Government data about Covid-19. The main result obtained from this work is to demonstrate the possibility of combining thick data with big data to achieve more comprehensive data analysis. Furthermore, among the main contributions of this research can be cited the artifact generated, that is, the model that adds thick data to a big data structure, with a focus on using qualitative data together with quantitative data Tue, 29 Jun 2021 00:00:00 GMT http://repositorio.jesuita.org.br/handle/UNISINOS/11164 2021-06-29T00:00:00Z http://repositorio.jesuita.org.br/handle/UNISINOS/11163 Predição de propriedades permoporosas de rochas carbonáticas utilizando redes neurais convolucionais Pasini, Augusto Thu, 01 Jul 2021 00:00:00 GMT http://repositorio.jesuita.org.br/handle/UNISINOS/11163 2021-07-01T00:00:00Z http://repositorio.jesuita.org.br/handle/UNISINOS/11053 DCARE: um modelo computacional para acompanhamento de pessoas com a doença de Alzheimer baseado em históricos de contextos e predições de contextos Machado, Savanna Denega The aging of the population generates the incidence of diseases characteristic of advancing age, among them Alzheimer’s desease. Patients with this disease, which affects brain functions, need support to maintain maximum independence and security during this stage of life, as the cure and reversal of symptoms have not yet been discovered. The daily monitoring technologies are an option tool to minimize the impacts caused in the daily lives of these people, ensuring greater patient safety and so that caregivers can monitor their activities, implementing a certain independence. In this context, this work aims to propose a model for monitoring patients with Alzheimer’s, seeking to synthesize the needs and characteristics that make up a better approach for its validation. The main scientific contribution of this work is the specification of a model that predicts user contexts for monitoring people with dementia during their daily lives, promoting accessibility to a tool for patient health and safety, in addition to contributing to the development of a datasets simulator with scenarios of daily activities of patients with Alzheimer’s disease. Based on the experimental research method, we sought to understand the disease and find solutions to minimize its impact on the daily monitoring of patients. To understand the problem, data on Alzheimer’s and the main difficulties faced by patients were collected through bibliographic research. From this information, the search for technologies that met the specified needs occurred. The functionalities employed were evaluated and points of improvement were identified. The project structure identifies the patient’s physiological data received from an external application, associating them with the model’s ontology, generating the context histories. Following the execution flow, Context Prediction techniques are used, which are based on the Context History data to generate prediction of future behaviors of patients, and perform a danger signal alert to the caregiver. The development of the scenarios used in the construction of the model were developed based on interviews with specialists in care for patients with Alzheimer’s disease. From the tests performed, with the mass of data generated by the developed simulator, called DCARE Dataset Simulator, the results of the predictions showed that the developed model reached the objective of the project, reaching 97.44% of general precision rate Fri, 04 Dec 2020 00:00:00 GMT http://repositorio.jesuita.org.br/handle/UNISINOS/11053 2020-12-04T00:00:00Z http://repositorio.jesuita.org.br/handle/UNISINOS/11052 A machine learning model for early diagnosis of arteriovenous fistula stenosis Hidalgo Junior, Orlando Vilmar Rodrigues Wed, 02 Dec 2020 00:00:00 GMT http://repositorio.jesuita.org.br/handle/UNISINOS/11052 2020-12-02T00:00:00Z http://repositorio.jesuita.org.br/handle/UNISINOS/11051 GPTR: um modelo para georreferenciamento de pessoas Santos, Matheus Hahn Dos Fri, 11 Dec 2020 00:00:00 GMT http://repositorio.jesuita.org.br/handle/UNISINOS/11051 2020-12-11T00:00:00Z http://repositorio.jesuita.org.br/handle/UNISINOS/11050 Uma arquitetura de orquestração de funções de rede de acesso de rádio desagregadas Renner, Julio The Fifth Generation (5G) Radio Access Network will be affected by significant changes. The usage of centralized resources and virtual network functions (VNFs) helps by reducing costs with the operation and acquisition of equipment by allowing the usage of generalpurpose hardware. However, the virtualization of network functions requires the development of new orchestration solutions or the customization of the existing ones in a way they meet the needs of a large and highly volatile environment with specific requirements. In today’s cloud environments, containers are widely used to deploy and virtualize applications. Several studies have already been done to provide VNFs using containers in the 5G scenario. In production environments, however, containers by themselves are not enough and orchestration tools need to be used in addition. The most widely adopted tool is Kubernetes, an open-source and highly customizable container orchestrator. In this context, this paper aims to analyze Kubernetes and its customization points for the usage in the orchestration of virtual network functions in the 5G networks radio access layer, proposing an architecture that aims to supply all the specific needs in this scenario following the standards defined by the standard development organizations. The results of the experiments applied demonstrated that K8S, through customizations, supports the requirements, for example, the positioning of the network functions using the RANPlacer and RANDeployer operators demonstrates an improvement of 230% compared with the native positioning of K8S Thu, 10 Dec 2020 00:00:00 GMT http://repositorio.jesuita.org.br/handle/UNISINOS/11050 2020-12-10T00:00:00Z http://repositorio.jesuita.org.br/handle/UNISINOS/11049 MALLi: um modelo de auxílio ao aprendizado de inglês por meio da música Gernhardt, Janaina Raquel More and more, nowadays, English becomes necessary as a second language. However, in the vast majority of times, learning a second language can be exhausting and it is possible to lose the motivation. Using something that is of the apprentice's taste is effective and music has been used to maintain that motivation. The growing use of mobile devices means that everything can be accessible from anywhere, making it possible to learn anytime, anywhere. The learning approach MALL (Mobile Assisted Language Learning) shows that it is possible to use the resources of mobile devices to help with language learning. In this context, this project aims to analyze how music can collaborate in learning English using MALL. For that, the research method DSR (Design Science Research) was used, which defines in stages how the research will be developed and evaluated. The MALLi model, a Mobile Assisted Language Learning for English, was developed by the present research. The model was evaluated and it was observed that it is possible to use a MALL with music to help in the challenges of learning English Thu, 10 Dec 2020 00:00:00 GMT http://repositorio.jesuita.org.br/handle/UNISINOS/11049 2020-12-10T00:00:00Z http://repositorio.jesuita.org.br/handle/UNISINOS/11048 ProFog: a proactive elasticity model for fog computing-based IoT applications Barth, Guilherme Gabriel Thu, 10 Dec 2020 00:00:00 GMT http://repositorio.jesuita.org.br/handle/UNISINOS/11048 2020-12-10T00:00:00Z http://repositorio.jesuita.org.br/handle/UNISINOS/11046 Detecção automática de casos de pneumonia por Covid-19 a partir de imagens de raio x do tórax e abordagens de Deep Learning Trombetta, Giordano Brunno Wagner Covid-19 has a high rate of transmission and contagion, and the early identification of new cases helps to prevent the transmission of the virus. This becomes a challenge, as the tests applied are usually done manually and take time. Initial studies have found cases of patients who have abnormalities on chest X-rays indicating the occurrence of this disease. With the study of these images, it is possible to create models with machine learning resources to automatically identify cases of COVID-19. In this study, experiments are carried out with convolutional neural network architectures to identify and evaluate a model of COVID-19 case detection with high precision in a fast and automatic way. Through the experiments carried out it is possible to affirm that the diagnosis by image of cases of severe acute respiratory syndrome based on X-ray exams is possible, having been observed results in which the model reaches 96 % accuracy when analyzing chest X-ray with three possible diagnoses in the experiments performed. The differential of this work in relation to the literature is the identification of multiple categories in the diagnosis, which is not observed in the studies studied Fri, 11 Dec 2020 00:00:00 GMT http://repositorio.jesuita.org.br/handle/UNISINOS/11046 2020-12-11T00:00:00Z http://repositorio.jesuita.org.br/handle/UNISINOS/11045 Proposição de uma metodologia pedagógica para a apresentação do algoritmo de Shor a iniciantes uma contribuição para o progresso da computação quântica sobre o algoritmo de criptografia RSA Cabral, Arthur Tassinari The Shor’s Algorithm is an important quantum algorithm that is present in the basis of many researches that deal with the impact of quantum computing on the RSA encryption algorithm. Through its use, and in an appropriate environment, this algorithm is able to solve the problem of prime factorization, putting the RSA algorithm at risk, which uses this problem to structure the public key. The learning about this algorithm can bring students closer to research involving the progress of quantum computing over RSA. However, it is noted that most of the resources available for the study of the Shor’s Algorithm involve, mainly, the reading of body texts, or models that approach the algorithm from a very technical perspective, without presenting a pedagogical method that proposes a better access to it. Given this scenario, the present article aims to promote a pedagogical methodology for the presentation of the Shor’s Algorithm to beginners. The methodology used is a web platform, which presents an educational step-bystep, with fixation exercises, providing user interactivity with the application. It was found that the platform was efficient and effective in introducing the Shor’s Algorithm to beginners, but that one can improve the approach on some of the steps of the algorithm. The article favors research that deals with the advancement of quantum computing over RSA because it is able to bring academia and people from the area of computing closer to them, and thus, consequently, benefits computer science Wed, 09 Dec 2020 00:00:00 GMT http://repositorio.jesuita.org.br/handle/UNISINOS/11045 2020-12-09T00:00:00Z http://repositorio.jesuita.org.br/handle/UNISINOS/11044 KATIE: modelo de tecnologia assistiva aplicando visão computacional para auxiliar pessoas com deficiência visual na identificação de medicamentos Steffenon, Jaqueline Dahmer Assistive technology aims to promote the inclusion of people with disabilities in daily activities through the use of technological innovations. razil has approximately 1 million blind people, so it is essential to develop technologies to improve the lives of this population. From the objective of promoting the autonomy of these people, the theme of this work is reached: the expansion of access to information to those who are deprived of it due to blindness. Routine tasks, such as identifying a drug pack or reading the packaging of a food product, become challenges for those who live with this disability. Routine tasks, such as identifying a drug pack or reading the packaging of a food product, become challenges for those who live with this disability. Inspired by existing solution Be my eyes, this work emerges in order to develop a technological solution to the problem and make these totally independent mundane tasks of others. In this sense, the KATIE model is presented, in which the user can ask a question about the scenario in which he finds himself and have an answer in real time. The results obtained with tests in the identification of medicines and cash notes, show that the prototype was able to fulfill the estimated objective, performing the recognition of the question and image data and returning an adequate answer to the user Sat, 11 Jul 2020 00:00:00 GMT http://repositorio.jesuita.org.br/handle/UNISINOS/11044 2020-07-11T00:00:00Z http://repositorio.jesuita.org.br/handle/UNISINOS/11042 Desenvolvimento de sensores inteligentes aplicados para a manutenção preditiva na indústria 4.0 Bonemberger Junior, Sérgio Luiz The fourth industrial revolution, also called Industry 4.0 is increasingly integrating the physical world with the digital world through the use of cyberphysical systems in industrial production plants. The development of this type of system uses the Internet of Things to enable equipment to exchange information with each other and their surroundings, thus constituting an intelligent industrial environment. Among many concepts that are emerging due to the fourth industrial revolution, one that can be highlighted is Predictive Maintenance, which seeks to predict when a machine will fail based on signals previously emitted by it. Considering the relevance of Predictive Maintenance using IoT sensors in the context of Industry 4.0, this work carried out a research in the areas of IoT and Fog Computing, which resulted in the development of a low cost architecture to collect data related to equipment operation in industries. Sensors were developed to collect machine vibration data, with configurable measurement periodicity, allowing them to maximize machine lifetime while maintaining the behavior of the collected data. To validate the proposed model, tests were performed with and without the use of the Fog Layer, with the use of different frequencies in the data collection and with or without the application of a data compression algorithm. The results did not prove that the use of a Fog Layer had a significant impact on reducing energy consumption. However, in tests with different frequencies in data collection, a relationship could be verified between the amount of data transmitted with the energy spent on operations Mon, 09 Dec 2019 00:00:00 GMT http://repositorio.jesuita.org.br/handle/UNISINOS/11042 2019-12-09T00:00:00Z http://repositorio.jesuita.org.br/handle/UNISINOS/11041 Parallel Monte Carlo tree search in general video game playing Centeleghe, Luis Gustavo Simioni Fri, 13 Dec 2019 00:00:00 GMT http://repositorio.jesuita.org.br/handle/UNISINOS/11041 2019-12-13T00:00:00Z http://repositorio.jesuita.org.br/handle/UNISINOS/11040 Reinforcement learning-based traffic lights controller with adaptive reward function Varisco, Fabio Müller O congestionamento no tráfego tem impactos na economia, na sustentabilidade das cidades e no bem-estar dos cidadãos. Este problema pode ser reduzido usando estratégias de controle de tráfego inteligentes para promover um uso mais eficiente da rede rodoviária. Controladores de semáforo baseados na técnica de Aprendizado por Reforço oferecem muitos benefícios em relação à outras técnicas, e um deles é ser capaz de ajustar as ações do controlador customizando a função de recompensa utilizada. Neste artigo, funções de recompensa são avaliadas sob diferentes condições de demanda e propomos uma função de recompensa adaptativa, que adapta seu objetivo dinamicamente de acordo com os níveis de saturação das vias Mon, 09 Dec 2019 00:00:00 GMT http://repositorio.jesuita.org.br/handle/UNISINOS/11040 2019-12-09T00:00:00Z http://repositorio.jesuita.org.br/handle/UNISINOS/11039 Otto: descobrindo assuntos e sentimentos em comentários Silva, Diego de Souza Tue, 10 Dec 2019 00:00:00 GMT http://repositorio.jesuita.org.br/handle/UNISINOS/11039 2019-12-10T00:00:00Z http://repositorio.jesuita.org.br/handle/UNISINOS/11038 Têmis: uma proposta de modelo computacional para a assistência no desaparecimento de pessoas Favretto, Aline Noronha According to the Brazilian Public Security Yearbook, the number of missing persons increased from 71,796 in 2016 to 82,684 in 2017. In addition, a lack of standard was also observed in the information used to register these missing persons. Given this scenario it is understood that a computer model based on artificial intelligence can assist in the registration and search of missing persons. Thus, the present work aimed to propose a computational model that assists in the search for missing persons in Brazil. This study was conducted through Design Science Research (DSR) whose goal is to contribute to both theory and practice. The evaluation of the model has shown potential mainly regarding support in the registration of missing persons. The proposed computational model applied the missing person register, also proposed in the present work, and obtained positive evaluations, besides raising evolution points in the form of suggestion of future works such as the use of facial recognition Thu, 12 Dec 2019 00:00:00 GMT http://repositorio.jesuita.org.br/handle/UNISINOS/11038 2019-12-12T00:00:00Z http://repositorio.jesuita.org.br/handle/UNISINOS/11037 Previsão de receita para empresas com o uso de diferentes técnicas de Aprendizado de Máquina Machado, Fernando Wed, 12 Dec 2018 00:00:00 GMT http://repositorio.jesuita.org.br/handle/UNISINOS/11037 2018-12-12T00:00:00Z http://repositorio.jesuita.org.br/handle/UNISINOS/11035 Moocare: utilizando predição de dados e Internet das Coisas para o gerenciamento da alimentação e produção de leite na pecuária leiteira Deon, Cássio Thu, 12 Jul 2018 00:00:00 GMT http://repositorio.jesuita.org.br/handle/UNISINOS/11035 2018-07-12T00:00:00Z http://repositorio.jesuita.org.br/handle/UNISINOS/11034 Migração para microserviços: uma metodologia de anteposição de serviços com indicadores de qualidade de software Colombo, Alexandre Brentano Mon, 09 Jul 2018 00:00:00 GMT http://repositorio.jesuita.org.br/handle/UNISINOS/11034 2018-07-09T00:00:00Z