Repositorio da Universidade da Coruña
Not a member yet
35055 research outputs found
Sort by
Sexting among adolescents : the role of Dark Triad in its prevalence and severity depending on sex
Full text link[Abstract] Sexting is an increasingly common form of communication among adolescents. Although this is not in itself a reprehensible or harmful act, it can become dangerous if this sharing is done under pressure, in a non-consensual manner, or has negative consequences for one of the parties involved. For this reason, this study aims to explore in depth the characteristics of the Dark Triad that may be related to the distribution of intimate images in the adolescent population and to investigate the possible differences in this relationship based on sex. Data were collected from a sample of 1570 adolescents aged 13 to 18. The results show a greater sexting behavior by girls, especially when it is carried out through social networks. Similarly, narcissistic and psychopathic traits are the traits most strongly associated with the presence and severity of image-sharing behavior in both boys and girls, with their importance varying according to the medium in which the image exchange takes place. Only Machiavellian traits seem to be associated with the occurrence of sexting among girls when it is done via mobile phones. In short, dark personality traits are relevant factors in the presence and severity of image-sharing behavior among adolescents, with differences observed according to the sex of the adolescents and the technological medium used
Effect of gender on patients with non-small-cell lung cancer treated with immune checkpoint inhibitors: a real-world study
Full text link[Abstract] Objective: To evaluate the differences in overall survival (OS) and progression-free survival (PFS) between men and women with non-small-cell lung cancer (NSCLC) treated with immune checkpoint inhibitors (ICIs) in second-line and later treatments. Methods: A retrospective, single-center observational study was conducted on patients with advanced NSCLC treated with ICIs (nivolumab, pembrolizumab, and atezolizumab) from January 2015 to December 2019 (with follow-up until December 2021). Demographic, clinical, and treatment-related variables were collected. OSand PFSwere analyzed using the Kaplan–Meier method and compared between genders using the log-rank test.A multivariate Cox regression analysis was performed to adjust for confounders. Results: A total of 189 patients were included, and 47 (25%) were women. The most common histology was adenocarcinoma (61%). Women began treatment at a younger age (59.8 vs. 66 years, p = 25 kg/m2, median OS was higher in women than in men (14.7 months vs. 10.1 months). Women had also a significantly worse PFS than men among those with a cumulative tobacco consumption of = 2 was significantly associated with both worse OS (HR = 3.53; 95% CI = 1.93–6.47) and PFS (HR = 2.19; 95% CI = 1.23–3.89). Women who discontinued due to toxicity (n = 7) had a median OS of 41.4 months (95% CI: 14.7–68.1) after discontinuation, whereas men (n = 15) had a median OS of 8.8 months (95% CI: 6.9–10.8), (p = 0.045). Conclusions: Sex-based differences were observed in the ICI outcomes. Women had worse PFS, particularly with lower BMI and lower tobacco exposure, despite similar OS between sexes. Women discontinued ICIs due to toxicity earlier but showed longer OS after discontinuation. Poor ECOG status was linked to worse outcomes across all the patients
Disentangling stellar atmospheric parameters in astronomical spectra using generative adversarial neural networks. Application to Gaia/RVS parameterisation
Full text link[Abstract] Context. The rapid expansion of large-scale spectroscopic surveys has highlighted the need to use automatic methods to extract information about the properties of stars with the greatest efficiency and accuracy, and also to optimise the use of computational resources.
Aims. We developed a method based on generative adversarial networks (GANs) to disentangle the physical (effective temperature and gravity) and chemical (metallicity and overabundance of α elements with respect to iron) atmospheric properties in astronomical spectra. Using a projection of the stellar spectra, commonly called latent space, in which the contribution due to one or several main stellar physicochemical properties is minimised while others are enhanced, it was possible to maximise the information related to certain properties. This could then be extracted using artificial neural networks (ANNs) as regressors, with a higher accuracy than a reference method based on the use of ANNs that had been trained with the original spectra.
Methods. Our model utilises auto-encoders, comprising two ANNs: an encoder and a decoder that transform input data into a low-dimensional representation known as latent space. It also uses discriminators, which are additional neural networks aimed at transforming the traditional auto-encoder training into an adversarial approach. This is done to reinforce the astrophysical parameters or disentangle them from the latent space. We describe our Generative Adversarial Networks for Disentangling and Learning Framework (GANDALF) tool in this article. It was developed to define, train, and test our GAN model with a web framework to show visually how the disentangling algorithm works. It is open to the community in Github.
Results. We demonstrate the performance of our approach for retrieving atmospheric stellar properties from spectra using Gaia Radial Velocity Spectrograph (RVS) data from DR3. We used a data-driven perspective and obtained very competitive values, all within the literature errors, and with the advantage of an important dimensionality reduction of the data to be processed.Acknowledgements. Horizon Europe funded this research [HORIZON-CL4-2023-SPACE-01-71] SPACIOUS project, Grant Agreement no. 101135205, the Spanish Ministry of Science MCIN/AEI/10.13039/501100011033, and the European Union FEDER through the coordinated grant PID2021-122842OB-C22. We also acknowledge support from the Xunta de Galicia and the European Union (FEDER Galicia 2021–2027 Program) Ref. ED431B 2024/21, CITIC ED431G 2023/01, and the European Social Fund – ESF scholarship ED481A2019/155Xunta de Galicia; ED431B 2024/21Xunta de Galicia; ED431G 2023/01Xunta de Galicia; ED481A2019/15
Breaking boundaries: Low-precision conditional mutual information for efficient feature selection
No full text©2025 Elsevier B.V. All rights reserved. This manuscript version is made available under the CC-BY-NC-ND 4.0 license https://creativecommons.org/licenses/bync-nd/4.0/. This version of the article has been accepted for publication in Pattern Recognition. The Version of Record is available online at https://doi.org/10.1016/j.patcog.2025.111375[Abstract]: As internet-of-things (IoT) devices proliferate, the need for efficient data processing at the network edge becomes increasingly critical due to the vast amounts of data generated. This paper presents a groundbreaking approach that leverages edge computing to address these challenges, using low-precision conditional mutual information (CMI) for feature selection. Our novel methodology improves the efficiency of edge computing systems by significantly reducing memory and energy consumption while maintaining high accuracy. We adapt this approach to feature selection algorithms, specifically, conditional mutual information maximization (CMIM) and incremental association Markov blanket (IAMB), and demonstrate its effectiveness for diverse datasets, including complex DNA microarrays. Our results show that low-precision methods not only compare competitively with traditional 64-bit implementations, but also yield significant performance and resource savings. For IoT and other machine learning applications, this work represents a significant advance in the development of more sustainable and efficient algorithms that can optimize computational resources and reduce their environmental impact.This work was supported in part by Xunta de Galicia/FEDER-UE under Grant ED431G 2023/01; Ministerio de Ciencia e Innovacion MCIN/AEI/10.13039/501100011033 and ‘Next-GenerationEU’/PRTR under Grants [PID2019-109238GB-C22; TED2021-130599A-I00; PID2023-147404OB-I00] and Ministry for Digital Transformation and Civil Service and ‘Next-GenerationEU’/PRTR under grant TSI-100925-2023-1. CITIC, as a Research Center of the University System of Galicia, is funded by Consellería de Educacion, Universidade e Formación Profesional of the Xunta de Galicia, Spain through the European Regional Development Fund (ERDF/FEDER) and the Secretaría Xeral de Universidades (ED431G 2023/01).Xunta de Galicia; ED431G 2023/0
Temporal Word Embeddings for Early Detection of Psychological Disorders on Social Media
Full text link[Abstract]: Mental health disorders represent a public health challenge, where early detection is critical to mitigating adverse outcomes for individuals and society. The study of language and behavior is a pivotal component in mental health research, and the content from social media platforms serves as a valuable tool for identifying signs of mental health risks. This paper presents a novel framework leveraging temporal word embeddings to capture linguistic changes over time. We specifically aim at at identifying emerging psychological concerns on social media. By adapting temporal word representations, our approach quantifies shifts in language use that may signal mental health risks. To that end, we implement two alternative temporal word embedding models to detect linguistic variations and exploit these variations to train early detection classifiers. Our experiments, conducted on 18 datasets from the eRisk initiative (covering signs of conditions such as depression, anorexia, and self-harm), show that simple models focusing exclusively on temporal word usage patterns achieve competitive performance compared to state-of-the-art systems. Additionally, we perform a word-level analysis to understand the evolution of key terms among positive and control users. These findings underscore the potential of time-sensitive word models in this domain, being a promising avenue for future research in mental health surveillance.The second and third authors thank the financial support supplied from projects: PLEC2021-007662 (MCIN/AEI/10.13039/501100011033 Ministerio de Ciencia e Innovación, European Union NextGenerationEU/PRTR) and PID2022-137061OB-C21 (MCIN/AEI/10.13039/501100011033/, Ministerio de Ciencia e Innovación, ERDF A way of making Europe, by the European Union); Consellería de Educación, Universidade e Formación Profesional, Spain (accreditations 2019-2022 ED431G/01 and GPC ED431B 2022/33) and the European Regional Development Fund, which acknowledges the CITIC Research Center. The first and fourth authors thank the financial support supplied by the Agencia Estatal de Investigación (Spain) (PID2022-137061OB-C22; PLEC2021-007662 MCIN/AEI/10.13039/501100011033, Plan de Recuperación, Transformación y Resiliencia, Unión Europea-Next Generation EU), Consellería de Cultura, Educación, Formación Profesional e Universidades (Centro de investigación de Galicia accreditation 2024-2027 ED431G-2023/04 and Reference Competitive Group accreditation 2022-2025, ED431C 2022/19) and the European Union (European Regional Development Fund - ERDF). The fourth author thanks the support obtained from project SUBV23/00002 (Ministerio de Consumo, Subdirección General de Regulación del Juego).
Open Access funding provided thanks to the CRUE-CSIC agreement with Springer Nature.Xunta de Galicia; ED431G/01Xunta de Galicia; ED431B 2022/33Xunta de Galicia; ED431G-2023/04Xunta de Galicia; ED431C 2022/1
Caso Breogan
Full text link[Resumen] El presente trabajo se adentra en el análisis de un caso práctico, el "Caso Breogán", que plantea cuestiones fundamentales relacionadas con la responsabilidad de los agentes de la edificación y el alcance de las pólizas de seguro de responsabilidad civil en el sector de la construcción. Este caso se centra en un proyecto de rehabilitación y ampliación de una vivienda unifamiliar en la provincia de A Coruña, donde se han identificado problemas relacionados con la cimentación y la omisión de estudios geotécnicos. Tales deficiencias han generado daños estructurales que derivan en la necesidad de evaluar las responsabilidades legales de los profesionales involucrados y de analizar el papel de las pólizas de seguro en la cobertura de estos daños.Traballo fin de grao (UDC.DER). Dereito. Curso 2024/202
Personalización, explicabilidad y sostenibilidad en sistemas de recomendación
Full text link[Resumen]: Este trabajo aborda la problemática de la transparencia en los Sistemas de Recomendación
(SR), proponiendo técnicas innovadoras para mejorar la explicabilidad visual, con un
enfoque en la sostenibilidad. Para ello, hemos llevado a cabo un estudio con tres modelos del
Estado del Arte (ELVis, MF-ELVis y BRIE) aplicados a reseñas de Tripadvisor. Las técnicas
empleadas se centran en optimizar la calidad de los datos de entrada, a través de tres estrategias
principales: 1) Selección de nuevos positivos y negativos con Aprendizaje Positivo y
Sin Etiquetas, 2) Aumentación de datos por transformación de imágenes o IA Generativa, y 3)
Mejora de embeddings de imágenes. Los resultados demuestran que la mejora de embeddings
incrementa el rendimiento en un 30%, reduce el tiempo de entrenamiento, y disminuye las
emisiones y el consumo en un 20% y 15% en entrenamiento e inferencia, respectivamente. La
combinación con las demás técnicas genera una mejora adicional del 5% en el rendimiento y
reduce significativamente las épocas de entrenamiento, disminuyendo las emisiones hasta un
60% adicional. Finalmente, estas técnicas han sido incluidas en una librería pública de carácter
modular, permitiendo su utilización en entornos de producción, así como en investigaciones
incrementales que partan de este trabajo.[Abstract]: This work addresses the issue of transparency in Recommender Systems (RS), proposing
innovative techniques to improve visual explainability, with a focus on sustainability. To this
end, we conducted a study with three state-of-the-art models (ELVis, MF-ELVis, and BRIE)
applied to Tripadvisor reviews. The employed techniques focus on optimizing the quality
of input data through three main strategies: 1) Selection of new positives and negatives using
Positive and Unlabeled Learning, 2) Data augmentation through image transformation or
Generative AI, and 3) Enhancement of image embeddings. The results show that improving
embeddings increases performance by 30%, reduces training time, and decreases emissions
and consumption by 20% and 15% in training and inference, respectively. The combination
with the other techniques provides an additional 5% performance improvement and significantly
reduces training epochs, lowering emissions by up to an additional 60%. Finally, these
techniques have been included in a publicly available modular library, allowing their use in
production environments as well as in incremental research building upon this work.Traballo fin de mestrado (UDC.FIC). Enxeñaría Informática. Curso 2024/202
Strategies for modelling roofs on large-scale urban drainage models focusing on incomplete data scenarios
No full text©2025. This manuscript version is made available under the CC-BY-NC-ND license. This is an accepted version of the following published document: https://doi.org/10.1016/j.uclim.2025.102362[Abstract:] Impermeable surfaces such as roofs play a key role in urban pluvial floods due to the rapid transfer of rainfall to drainage networks, contributing to system overloading. This study proposes and evaluates different modelling strategies within large-scale urban drainage models, exploring simplified approaches for roof geometry and roof-to-manhole connections. The results indicate that, regardless of the methodology used to estimate roof width, the differences become negligible if the discharge point is distant. Additionally, for a contributing area of roofs discharging upstream to a manhole, the method of roof-to-manhole connection does not have a significant influence, which demonstrates the potential of these strategies to streamline the modelling process without compromising the reliability of the simulations. The findings highlight the feasibility of applying these modelling strategies in situations where data completeness is not feasible, offering a balanced solution between model complexity and accuracy.This project has received financial support from the Spanish Ministry of Science and Innovation (MCIN/AEI/10.13039/501100011033) within the project “SATURNO: Early warning against pluvial flooding in urban areas” (PID2020-118368RB-I00) and the FPI predoctoral grant from the Spanish Ministry of Science, Innovation, and Universities (PRE2021-098425). The contract of Esteban Sañudo is funded by the project “DRAIN - Digital RAIN. An Integrated Urban Drainage Model” (CPP2021-008756) funded by MICIU/AEI/ 10.13039/501100011033 and by the “European Union NextGenerationEU/PRTR” and the author also acknowledges the support of the INDITEX-UDC 2021 and 2022 Predoctoral Grants
Aplicación Web Java para la gestión de reservas en pisos u hoteles
No full text[Resumen]: En la actualidad, el turismo vacacional está en constante crecimiento impulsado por una mayor accesibilidad a las opciones de transporte y alojamiento; gracias a este crecimiento se ha experimentado un incremento en la demanda de plataformas que centralicen la búsqueda y reserva de alojamientos.
Por ello, este trabajo que llevará por nombre DeepDive, tiene como objetivo el análisis, diseño e implementación de una aplicación web para gestionar ofertas de alojamiento en un mismo sitio web. La aplicación permitirá consultar las diferentes ofertas de alojamiento en cualquier ubicación geográfica y también permitirá realizar reservas, valorarlas y subir tu propio alojamiento. El principal objetivo de DeepDive es proporcionar a los usuarios una herramienta que les permita de manera rápida, sencilla e intuitiva consultar, buscar y reservar alojamientos en cualquier ubicación geográfica. También permitirá a un usuario publicar su alojamiento y actualizarlo siempre que lo considere conveniente. DeepDive también proporcionará varias vistas específicas para consultar tanto tus reservas como tus alojamientos. Una vez se ha cumplido el periodo de una reserva, y con el fin de promover la calidad del servicio, los usuarios
podrán escribir una valoración sobre su experiencia en dicho alojamiento. DeepDive busca que la aplicación sea inclusiva, integrando opciones de accesibilidad para que los usuarios puedan personalizar la interfaz a sus necesidades visuales. Además la aplicación está internacionalizada en varios idiomas. La aplicación integra un servicio web real que permite acceder a una API con una amplia variedad de alojamientos, lo que permitirá a los usuarios acceder a una mayor variedad de resultados y opciones al buscar alojamientos, lo que facilita que las búsquedas sean más completas y precisas, optimizando el tiempo de los usuarios y facilitando la selección de su alojamiento ideal. La aplicación se ha desarrollado utilizando Maven, Java y Spring para la arquitectura backend y React, Node, JavaScript y CSS para la arquitectura frontend. Para el desarrollo del proyecto se ha empleado la metodología ágil Scrum, dividiendo el trabajo en tareas manejables que se ejecutarán de manera iterativa en distintos Sprints; una vez finalizado el Sprint se validará que las funcionalidades implementadas cumplen con los requisitos establecidos en la planificación.[Abstract]: Currently, vacation tourism is constantly growing, driven by greater accessibility to transportation and lodging options; as a result of this growth, there has been an increase in demand for platforms that centralize the search and booking of lodges. Therefore, this work, named DeepDive, aims to analyze, design, and implement a web application to manage lodge offers on a single website. The application will allow users to search for different lodge offers in any geographic location, make bookings, rate them, and upload their own lodges. The main objective of DeepDive is to provide users with a tool that allows them to quickly, easily, and intuitively search, browse, and book lodges in any geographic location. It will also allow a user to publish their lodge and update it whenever they deem appropriate. DeepDive will also provide several specific views to consult both your bookings and your lodges. Once a booking period has ended, and in order to promote service quality, users will be able to write a review about their experience at the lodge. DeepDive aims for the application to be inclusive, integrating accessibility options so that users can customize the interface to their visual needs. Additionally, the application is internationalized in multiple languages. The application integrates a real web service that allows access to an API with a wide variety of lodges, enabling users to access a greater variety of results and options when searching for lodges, which makes searches more complete and accurate, optimizing users’ time and
facilitating the selection of their ideal lodge. The application has been developed using Maven, Java, and Spring for the backend architecture, and React, Node, JavaScript, and CSS for the frontend architecture. For the project’s development, the agile Scrum methodology has been employed, dividing the work into manageable tasks that will be executed iteratively in different Sprints; once the Sprint is completed, it will be validated that the implemented features meet the requirements set in the planning.Traballo fin de grao (UDC.FIC). Enxeñaría Informática. Curso 2024/202
Job satisfaction and violence in the clinical relationship in physical therapists in Spain
Full text link[Resumen] Objetivo. Conocer la satisfacción laboral (SL) de los/las fisioterapeutas en España según el efecto de la violencia ocupacional y otros factores sociodemográficos, de salud y laborales.
Diseño. Estudio transversal.
Emplazamiento. Todos los niveles de atención (atención primaria, comunitaria y hospitalaria) y sector público y privado en España.
Participantes. Fisioterapeutas que han trabajado al menos 3 meses en el último año con respuesta completa a las variables de interés (n = 2.590).
Mediciones principales. Valoración de la SL, variables sociodemográficas, laborales, del estado de salud y sintomatología y de padecimiento de violencia laboral. Se realizó un análisis descriptivo y tres modelos de regresión logística.
Resultados. La SL media de los/las fisioterapeutas es de 7,26 puntos, llegando a 8, o más, en el 46,8% de los casos, con mayor porcentaje en hombres. Referir no haber padecido violencia psicológica se relacionó con una mayor probabilidad de tener SL, incluso controlando por el resto de las variables estudiadas (OR1 = 0,485; OR2 = 0,611; OR3 = 0,697, respectivamente, para cada modelo).
Las variables vinculadas a la salud (estado de salud, síntomas, consumo de tabaco/alcohol/otras sustancias) y al ámbito laboral (jornada, área laboral, autonomía, relación con superiores/compañeros/as) se relacionaron de forma estadísticamente significativa con la SL.
Conclusiones. Casi el 47% de las personas encuestadas presentaron valores de SL muy elevados. Ciertas áreas de trabajo, así como factores positivos de la salud, se han vinculado con una muy alta SL. La violencia psicológica es la forma de violencia que, independientemente de los otros factores analizados, conlleva una menor SL.[Abstract] Objective. To assess the job satisfaction (JS) of physiotherapists in Spain and their relationship with occupational violence, as with other socio-demographic, health, and occupational factors.
Design. A cross-sectional study was conducted.
Setting Primary, community, and hospital attention level at public and private care in Spain.
Participants. Physiotherapists in Spain who have been working for at least 3 months during the last year, and with complete answer to the required variables (n = 2,590).
Main measurements. Information was collected through a questionnaire distributed online. A descriptive quantitative analysis and 3 logistic regression models were performed. In the first model, sociodemographic and violence variables were included as independent variables, in the second, health-related variables, and in the third, occupational variables.
Results. The average JS of physiotherapists is 7.26 points, being 8 or more in 46.8% of the cases, with a higher percentage in men”. Referring not having suffered psychological violence was related to a higher probability of having JS, even controlling for the rest of the variables studied (OR1 = 0.485; OR2 = 0.611; OR3 = 0.697, respectively for each model).
Variables related to health (state of health, symptoms, consumption of tobacco/alcohol/other substances) and to the work environment (working day, work area, autonomy, relationship with superiors/colleagues) were statistically significantly related to JS.
Conclusions. Almost 47% of the respondents had very high JS values. Certain areas of work as well as positive health factors have been linked to very high JS. Psychological violence is the form of violence that, independently of the other factors analysed, leads to lower JS