7664 research outputs found
Sort by
Monitorização em Contínuo do Risco de Incêndio Florestal
Get PDFO aumento de ondas de calor e períodos de seca prolongados, provocados pelas
alterações climáticas, têm levado a um aumento significativo do risco de incêndio. É por
isso importante estudar metodologias que auxiliem na sua previsão por forma a
minimizar o número de ocorrências, a sua dimensão e consequências. Um dos
principais fatores que determinam o risco de incêndio florestal é o teor de humidade dos
combustíveis florestais, dado que influencia quase todos os fatores relacionados com o
desenvolvimento de um incêndio, no entanto, a sua determinação pode ser trabalhosa
e dispendiosa. Para tal, existem métodos expeditos que determinam o teor de humidade
de combustíveis florestais em campo, permitindo uma monitorização à distância com
menos custos e em tempo real. O objetivo central do presente trabalho foi o
desenvolvimento de métodos de medição de parâmetros ambientais e florestais que
permitam monitorizar em contínuo o risco de incêndio de uma determinada região, com
representatividade temporal e espacial e a um custo reduzido, com a utilização de
sensores de vareta (fuel stick), sensores de teor de humidade do solo e células de carga
para uma monitorização ininterrupta da variação do peso de uma amostra de
combustível florestal. No presente trabalho, foram testados vários métodos de
determinação de teor de humidade de combustíveis, nomeadamente com um fuel stick,
com sensores de humidade do solo e com uma balança com duas células de carga.
Como forma de verificar a fiabilidade de um dos métodos testados, nomeadamente o
fuel stick, os resultados da medição do teor de humidade do fuel stick foram comparados
com índices de perigo de incêndio horários e diários, com o número de ocorrência de
incêndios e áreas ardidas e com o teor de humidade modelado. Verificou-se que o teor
de humidade do fuel stick se relaciona significativamente com os índices de perigo de
incêndio FFMC, ISI e FWI, especialmente nos meses de julho e agosto, quer se tratem
de valores horários ou diários. Da relação com o índice FWI resultou uma equação de
regressão que possibilita definir as classes de risco de incêndio. Considerando os dados
dos incêndios, resultou uma relação significativa entre o teor de humidade do fuel stick
com o número de ocorrências diárias. Daqui resultou uma equação que permite estimar
o número de ocorrências diárias com base no teor de humidade do fuel stick registado
às 12 UTC. Relativamente à relação com o teor de humidade modelado, concluiu-se
que o melhor modelo é o que utiliza ramos de pinheiro-bravo morto, de 3 cm a 5 cm de
diâmetro, classificados como combustíveis de 100h e que os valores do teor de
humidade do fuel stick são sempre mais reduzidos do que o teor de humidade
modelado.ABSTRACT:
The increase in heat waves and prolonged periods of drought caused by climate change
has led to a significant increase in the risk of fire. It is therefore important to study
methodologies that help predict them to minimize the number of occurrences, their size
and consequences. One of the main factors that determines the risk of forest fires is the
moisture content of forest fuels, as it influences almost all factors related to the
development of a fire. To this end, there are expedient methods for determining the
moisture content of forest fuels in the field, allowing for less costly remote monitoring in
real time. The main objective of this work was to develop methods for measuring
environmental and forestry parameters that allow continuous monitoring of the fire risk
in each region, with temporal and spatial representativeness and at a low cost, using fuel
stick sensors, soil moisture content sensors and load cells for uninterrupted monitoring
of the variation in the weight of a forest fuel sample. In this work, various methods of
determining the moisture content of fuels were tested, namely with a fuel stick, soil
moisture sensors and a scale with two load cells. To verify the reliability of one of the
methods tested, namely the fuel stick, the results of measuring the moisture content of
the fuel stick were compared with hourly and daily fire danger indices, with the number
of fires and burnt areas and with the modeled moisture content. The moisture content of
the fuel stick was found to be significantly related to the FFMC, ISI and FWI fire danger
indices, especially in the months of July and August, both for hourly and daily values.
The relationship with the FWI index resulted in a regression equation that makes it
possible to define fire risk classes. Considering the fire data, there was a significant
relationship between the moisture content of the fuel stick and the number of daily
occurrences. This resulted in an equation that makes it possible to estimate the number
of daily occurrences based on the moisture content of the fuel stick recorded at 12 UTC.
Regarding the relationship with the modeled moisture content, it was concluded that the
best model is the one that uses dead maritime pine branches, from 3 cm to 5 cm in
diameter, classified as 100h fuels, the modelling parameters for 100h fuels and that the
values of the moisture content of the fuel stick are always lower than the modeled
moisture content
Viabilidade para adoção da tecnologia blockhain na gestão de contratos e de dados na construção civil integrado ao BIM
Get PDFConsiderando a escassa aplicação de novas tecnologias na área da Engenharia Civil,
que possam ajudar a melhorar os processos de registos de dados, auditabilidade de etapas da
construção e aperfeiçoar a qualidade dos serviços entregues, nomeadamente, desenvolvimento
de projetos, gestão do canteiro de obras, cronogramas etc.., a tecnologia blockchain é trazida à
luz da comunidade científica para evidenciar as possibilidades da sua utilização. Trata-se de
uma ferramenta para auxiliar os processos de gestão na construção civil. Verifica-se uma
crescente procura por opções que possam trazer mais transparência e confiabilidade entre as
partes interessadas nos processos de gestão da construção. O trabalho tem assim como
principal objetivo analisar de que forma a tecnologia blockchain pode contribuir na gestão de
contratos integrados com Building Information Modelling (BIM) e responder à pergunta de
investigação seguinte:
Como pode a tecnologia blockchain contribuir no contexto de segurança dos dados na
indústria da construção civil?
Para responder à questão realiza-se uma análise comparativa à revisão bibliográfica e
o estudo de caso de uma empresa na área da construção civil que trabalha com a tecnologia
blockchain, possibilitando o confronto de informações através da análise de dados primários,
obtidos pela aplicação de questionário no estudo de caso e dados secundários, através da
investigação da literatura. Desse modo os resultados são analisados, comparados e
comentados através da triangulação dos dados, oferecendo um panorama das fontes de
evidencias permitindo concluir que as características da tecnologia blockchain podem ser
uteis e viáveis no setor da construção civil.ABSTRACT:
Considering the scarce application of new technologies in the area of Civil
Engineering, which can help improve data recording processes, auditability of construction
stages and improve the quality of services delivered, namely project development,
construction site management, timelines, etc., blockchain technology is brought to the
attention of the scientific community to highlight the possibilities of its use. It is a tool to
assist management processes in construction. There is a growing search for options that can
bring more transparency and reliability among interested parties in construction management
processes. The main objective of this work is to analyze the benefits of using blockchain
technology in the management of contracts integrated with Building Information Modeling
(BIM) and answer the following research question:
How can blockchain technology contribute in the context of the construction industry?
To answer the question, a comparative analysis is carried out with the bibliographical
review and the case study of a company in the construction sector that works with blockchain
technology, enabling the comparison of information through the analysis of primary data,
obtained by applying a questionnaire. in the case study and secondary data, through literature
investigation. In this way, the results are analyzed, compared and commented through data
triangulation, offering an overview of the sources of evidence, allowing us to conclude that
the characteristics of blockchain technology can be useful and viable in the construction
sector
The image of Portugal as a destination for equestrian Tourism
Get PDFThe purpose of this work is mainly based on knowing equestrian tourism as
a tourism product and as a form of tourism in Portugal, this way, allowing the topic of
the image that Portugal holds, as an equestrian tourism destination, to develop. With
this in mind, it is intended to answer the following questions: 1) Who is the tourist who
practices equestrian tourism in Portugal? 2) What is the image that Portugal holds as an
equestrian tourism destination? 3) What are the motivations behind tourists who practice
this sport in Portugal?info:eu-repo/semantics/publishedVersio
NEW TECHNOLOGIES IN EDUCATION AND SOCIETY - BOOK OF ABSTRACTS
Get PDFNew technologies, like artificial intelligence and virtual collaborative environments, are in our daily lives
and are becoming part of the fabric of human interactions and thought processes. They are powerful
tools to improve our well-being and development that also challenge and create risks to our known
processes and ways of being in the world. In Education, and in other areas, several projects,
initiatives, research and general use offer material for critical analysis and planning of future avenues.
As in other complex challenges, networked discussions, learning from each other and creating
community is an important part of the solutions to be charted. Partners in the European University for
Customised Education (EUNICE) have combined efforts to organise the event Seminar New
Technologies in Education and Society (NTES), allowing deep discussions based on real and/or
research-based uses and cases across different countries and contexts.
This 3-day seminar, featuring plenary sessions, discussions, and poster sessions, will dive into the
exploration and discourse surrounding new technologies in education and society. The plenary
sessions will showcase new technologies utilized in education, while also serving broader social
needs. Attendees will gain insight into the innovative concept of the EUNICE Virtual Lab, explore the
opportunities and challenges posed by the Metaverse, Artificial Intelligence (AI), and blended
real/virtual teaching approaches. Attendees will also have the unique opportunity to experience the
virtual world first-hand through the use of VR headsets.
This seminar aligns with the EUNICE4U project, which aims to develop a shared system of support for
pedagogical innovation. One of the primary goals of EUNICE activities is to improve student learning
through continuous pedagogical innovation, including effective integration of learning technologies,
interdisciplinary teaching methodologies, and challenge-based learning approaches. EUNICE strives
to harness the potential of immersive learning, particularly through virtual collaborative environments,
as a promising method for delivering distance academic programmes.info:eu-repo/semantics/publishedVersio
Content Matching and Sentiment Analysis
Get PDFDeveloping new services or improving existing ones is becoming more accessible with
the evolution of Natural Language Processing (NLP) techniques. Chatbots are a known
example of an NLP-based service; they can interact with humans using text messages or
natural language. NLP grants, however, the development of other types of services based
on natural languages, such as machine translation, email spam detection, information
extraction, content summarization, and question answering. A current need, to develop
smart cities projects, is a system that can match content (text) from a project offer
description with the candidates description by finding common patterns in different textual
descriptions. This project presents an implementation of an automated tool with AI and
NLP to match needs and concrete ideas for innovation with the skills and offers of the
business sector, including start-ups and entrepreneurs. In sentiment analysis, NLP can be
harnessed to recognize and categorize the emotional tone conveyed in textual content, such
as project collaborator reviews, customer reviews, or social media posts. The sentiment
analysis component in this project establishes a tool for comprehending and categorizing
sentiments, for candidates seeking engagement in smart cities projects
Unlocking portfolio resilient and persistent risk: A holistic approach to unveiling potential grounds
No full textPurpose: This study identifies residual, persistent, or resilient risks that remain even after controls
for systematic and sentiment risk and extensive portfolio diversification are applied.
Method: This methodology employs the Newey-West regression analysis in conjunction with the
Lagrange multiplier method to construct a global minimum variance portfolio. This analysis focuses explicitly on the diversifiable risk component. It uses two benchmarks, the Wilshire 5000
and S&P500, in collaboration with investor sentiment metrics. Its primary objective is to mitigate
systematic and idiosyncratic risk by examining three different portfolios (Tourism, Utilities/Energy, and Industrials) comprising 132 individual stocks observed over a span of 17 years.
Findings: This study identifies a persistent and resilient residual risk that may be connected to
undisclosed uncertainties and emerging risks that are known to exist but are not yet fully
materialized. These include potential ramifications from emerging widespread climate disasters,
the duration of the recession periods, the effect of uncertainty on merger and acquisition outcomes, and even unknown threats such as the proliferation of new computer viruses and system
vulnerabilities, the repercussions of unregulated artificial intelligence, and shifts in individual
preferences, beliefs, and behaviours that may influence both investors and society’s economic
dynamics. All these risks and uncertainties will greatly contribute to a fear of the unknown and
subsequently affect financial markets.
Novelty: In this study, we extract the systematic and investor sentiment risks for individual socks
and construct annual minimum variance portfolios. Our next step was to justify the presence of
unknown, persistent, resilient, or residual risk factors.
Practical implications: This particular approach provides multiple advantages to investors and
regulators. It enables them to construct portfolios with lower levels of risk and proactively
mitigate potential sources of risk that are presently little more than possibilities but may evolve
into real threats.info:eu-repo/semantics/publishedVersio
Exploring E-portfolios: Illuminating Accounts of the Pedagogical Innovation Training Programme at the Polytechnic Institute of Viseu
Get PDFEvery educational institution strives for pedagogical excellence, driven by the goal of providing the most effective and impactful learning experiences to its students. This is no different at the Polytechnic Institute of Viseu (IPV) and other Polytechnic Institutes participating in a Pedagogical Innovation Training Programme developed within a consortium committed to enriching educational methodologies and tools. There is evidence that innovative pedagogical methodologies lead to enhanced student engagement, foster meaningful interactions, promote critical thinking and problem-solving skills, and ultimately better academic achievement. This study focuses on the training course on pedagogical innovation offered to the teaching staff from IPV and vocational schoolteachers from the region, by examining their reflective portfolios. We aim at illuminating the impact and efficacy of the initiative in fostering active methodologies and innovative pedagogical tools, employing qualitative analysis to uncover the nuanced perceptions of the IPV participants in the six editions of the programme (2021-2023). The findings reveal that they value active methodologies, intercultural and multidisciplinary collaboration, and the integration of industry- aligned skills development, even if we encounter accounts of challenges faced during the implementation process of the training course. Ultimately, this study contributes to assessing the initiative’s impact and underscores the pivotal role of innovative teaching methodologies in striving for educational excellence. In light of the findings, policy recommendations include encouraging continued investment in pedagogical innovation training programmes, supporting interdisciplinary collaboration, fostering industry alignment, and addressing implementation challenges to ensure the effectiveness of such initiatives.info:eu-repo/semantics/publishedVersio
Enhancing Interpretability of Neural Networks in Food Recommendation Systems
Get PDFABSTRACT:
Over the years the risk of developing diseases related to poor alimentation has
been increasing. Many of these diseases are caused by obesity. Obesity is a
silent disease related to being overweight, which due to its rapid growth has
become a public health problem. Worldwide obesity has nearly tripled since
1975. Obesity can lead to health problems like type 2 diabetes, cardiovascular
disease, and even cancer. The main factors that result in obesity are a sedentary
lifestyle and a poor diet. Although obesity is uncured, it can be avoided/treated
through a healthier lifestyle and diet. Amid so much information about diets
and healthier recipes, it can be difficult to find a diet that meets the needs
of each person. Recommendation systems can filter from a large dataset, the
information that best suits the profile of each user. Due to the constant in crease in information and computational power, recommendation systems have
evolved from a traditional approach to a deep-learning one. Recommendation
systems are a hot topic in deep learning. Research in the food recommendation
systems area has seen little development when compared to recommendations
systems in other areas, such as leisure and entertainment. A powerful tool to
use in food recommendation systems is neural networks. Neural networks play
an important role in our society, for their capacity to learn from complex and
high-dimensional data. One side down of neural networks is the difficulties if not
impossibility in understanding how the predictions are being made. The behind the-scenes often remain opaque, leading neural networks to be characterized as
“black boxes”. With this research, we aim to give contribute to understanding
how neural networks operate underneath and make them more transparent and
so more trustworthy. With this goal in mind, we propose the use of a secondary
model to predict the errors of a primary neural network. By analyzing the error
predictions of the second model, we aim to gain insights into its decision-making
process. With this approach, we hope not only to help to understand the func tioning of neural networks but also to provide an idea of how to improve their
performance. Improving neural networks’ understanding can make them more
simple and accessible. With the work developed through this research, we look
to stride towards making neural networks more transparent and explainable,
thereby enhancing trust in these powerful models.RESUMO:
Ao longo dos anos, o risco de desenvolver doenças relacionadas a má alimentação
tem aumentado. Muitas destas doenças são causadas pela obesidade. A obesidade é uma doença silenciosa relacionada com o excesso de peso, que devido
ao seu rápido crescimento tornou-se um problema de saúde pública. A nível
mundial a obesidade quase que triplicou desde 1975. A obesidade pode levar
a problemas de saúde como diabetes do tipo 2, doenças cardiovasculares e até
mesmo cancro. Os principais fatores que resultam na obesidade são um estilo de
vida sedentário e uma dieta pobre. Embora a obesidade não tenha cura, pode
ser evitada/tratada através da adoção de um estilo de vida e de uma dieta mais
saudáveis. No meio de tanta informação sobre dietas e receitas saudáveis, pode
ser difícil encontrar uma dieta que satisfaça as necessidades de cada pessoa. Os
sistemas de recomendação podem filtrar a partir de um grande conjunto de dados a informação que melhor se adapta ao perfil de cada utilizador. Devido ao
constante aumento de informação e de poder computacional, os sistemas de recomendação evoluíram desde uma abordagem tradicional para uma abordagem
de deep learning. Os sistemas de recomendação são um tema quente na ´area
de deep learning. A investigação na ´area dos sistemas de recomendação alimentar tem visto pouco desenvolvimento quando comparada com os sistemas de
recomendação em outras ´areas, como lazer e entretenimento. Uma ferramenta
poderosa a utilizar nos sistemas de recomendação alimentar são as redes neurais.
As redes neurais desempenham um papel importante na nossa sociedade, pela
sua capacidade de aprender a partir de dados complexos e de alta dimensão.
Um dos lados negativos das redes neurais é a dificuldade, se não a impossibilidade, de compreender como as previsões estão a ser feitas. Os processos de
decisão permanecem frequentemente opacos, levando as redes neurais a serem
caracterizadas como ”caixas pretas”. Com esta investigação, pretendemos contribuir para a compreensão de como as redes neurais operam debaixo dos panos e
torná-las-ás mais transparentes e, portanto, mais confiáveis. Com este objetivo em
mente, propomos o uso de um segundo modelo para prever os erros de uma rede
neural. Ao analisar as previsões de erro do segundo modelo, pretendemos obter
noções sobre o processo de tomada de decisão da rede neural. Com esta abordagem, esperamos não só ajudar a entender o funcionamento das redes neurais,
mas também fornecer uma ideia de como melhorar o seu desempenho. Melhorar a compreensão das redes neurais pode torná-las-ás mais simples e acessíveis. Com o trabalho desenvolvido através desta investigação, procuramos avançar no sentido de tornar as redes neurais mais transparentes e explicáveis, aumentando assim a confiança nestes modelos poderosos
Livro de Resumos do Simpósio Internacional em Educação Especial e Inclusiva (SIEEI) 2024
Get PDFO Simpósio Internacional em Educação Especial e Inclusiva (SIEEI): Investigação e Práticas destina-se a promover o debate sobre Educação Inclusiva e incentivar a partilha de saberes e reflexão sobre as práticas educativas na área, orientando-se para todos os que procuram dar respostas aos desafios da educação na atualidade. O evento enquadra-se no âmbito das atividades científicas do Mestrado em Educação Especial – Domínio Cognitivo e Motor, da Escola Superior de Educação de Viseu e conta com diversos contributos de profissionais e investigadores na área da educação especial e inclusiva, contemplando a possibilidade de apresentação de comunicações livres (orais e posters).Destaca-se a colaboração de oradores nacionais e internacionais: Jesus Molina, da Universidade de Múrcia (Espanha), Grupo de Investigación sobre Diversidad Funcional y Derechos Humanos, Anabela de Oliveira Carvalho (Agrupamento de Escolas Infante D. Henrique, Instituto Politécnico de Viseu, Escola superior de Tecnologia e Gestão), Maria de Fátima Almeida (Agrupamento de Escolas de Nelas, Universidade Católica Portuguesa, Faculdade de Educação e Psicologia), Marisa Carvalho (Universidade Católica Portuguesa), Susana Jimenez (Investigadora na Argentina, Espanha e Portugal), Sofia Campos (Instituto Politécnico de Viseu, Escola Superior de Saúde) e Jorge Humberto Dias (representante de Portugal no Projeto Happy Schools). Este livro de resumos procura sistematizar contributos relevantes no domínio da inclusão, destacando-se a diversidade de olhares sobre questões contemporâneas da educação e reabilitação de pessoas com necessidades educativas específicas.info:eu-repo/semantics/publishedVersio
Examining the role of familiarity in the destination word-of-mouth
Get PDFAfter the COVID-19 pandemic, tourism has grown enormously; this growth has been even more significant in Portugal. The dissemination of experiences by tourists who have already visited a region significantly impacts the interest aroused by other tourists. It is increasingly common for tourists to analyse the experiences of others before they travel. In this sense, word-of-mouth by tourists is a factor to be considered in tourism. Since tourism is based on experiences and familiarity with destinations is reinforced by accumulated experiences, studying the effects of familiarity on tourists' intention to word-of-mouth would be relevant. Furthermore, given that the attitude towards the region visited comprises cognitive and affective dimensions, it would be relevant to analyse the effects of the attitude towards the region on the intention to word-of-mouth.
In this sense, this study sought to analyse the determining factors of word-of-mouth in tourism. To this end, a quantitative, cross-sectional survey was carried out, for which data was collected through a questionnaire from a sample of 906 tourists using the Smart PLS software. The results showed that familiarity with the region and attitudes towards the region is essential in tourists' intention to WOM. Therefore, this study supports the idea that tourist destinations should welcome tourists in the best way possible by providing them with greater familiarity with the region so that, through the word-of-mouth generated, it is possible to attract more tourists.info:eu-repo/semantics/publishedVersio