Universidad de Los Andes
Repositorio institucional Séneca Universidad de los Andes (Colombia)Not a member yet
38709 research outputs found
Sort by
El ciclo del agua y las contribuciones de la naturaleza en América Latina y el Caribe Desafíos y oportunidades de transformación
No full textEste resumen de políticas ha sido elaborado como insumo para la XXIV Reunión del Foro de ministros de Ambiente de América Latina y el Caribe (ALC), de Lima, Perú, septiembre de 2025.
Este Policy Brief sintetiza seis mensajes clave para orientar la acción urgente de los tomadores de decisión de América Latina y el Caribe. La región enfrenta un momento decisivo: la pérdida de ecosistemas, el deterioro de servicios hídricos y el aumento de eventos extremos generan crecientes costos sociales y económicos, especialmente en comunidades rurales, indígenas y costeras. Este contexto crítico también abre oportunidades para transformar el modelo de desarrollo hacia uno más justo y sostenible mediante reformas estructurales y soluciones basadas en la naturaleza
Modelo de formación STEM para niños y jóvenes migrantes venezolanos
No full textThis dissertation proposes a DEI-informed, ICT-enhanced STEM education model for Venezuelan migrant children and youth in Colombia that both identifies what learners currently know, value, and believe about STEM and creates the conditions for those profiles to improve in ways that make STEM a plausible higher-education pathway. The model is built as an integrated, mixed-methods system with five interlocking layers. First, a context and equity layer addresses invisibilization and identity threats by explicitly valuing migrants’ funds of knowledge and lived experiences, positioning inclusion and belonging, not remediation, as the entry point to participation (Pain, 2004; Abrego, 2014; Claeys-Kulik et al., 2019; Dobscha, 2021; Zheng, 2022). Second, a teacher-capacity layer develops DEI-aligned practice through professional learning that emphasizes inclusive and active pedagogies, with teachers iteratively interpreting and enacting new routines in their own settings (Dewsbury et al., 2022; Castillo-Montoya et al., 2023; Kim et al., 2023; Zappe et al., 2022). Third, a learning-design layer specifies culturally relevant, project-based tasks mediated by ICT to widen access, personalize learning, and scaffold language and trauma-aware participation, operationalizing blended/b-learning affordances as equity tools rather than mere delivery modes (Galvis & Pedraza, 2013). Fourth, a learner-facing layer translates design principles into a hybrid Python course organized by “big ideas” and active methods so that students experience early mastery in decomposition, conditional reasoning, and iteration, dispositions central to computational thinking and long-term interest (Palmer, 2012; Galvis Panqueva, 2021). Fifth, a measurement-and-feedback layer uses a perceptions questionnaire anchored in public understanding of science indicators, knowledge, interest, and attitudes (KIA), to generate ethical, comparable baselines and monitor change, complemented by qualitative evidence from multiple case studies and focus groups (Miller, 1983, 2012; Durant, Evans, & Thomas, 1989; Durant, Bauer, & Gaskell, 1998; CONACYT, 2017; Minciencias y Observatorio de Ciencia y Tecnología, 2015; ANII, 2014; SENACYT, 2017; Stake, 1998; Yin, 2009). The overall architecture is underpinned by a PRISMA-based review that justifies construct selection and design choices and by a pragmatic mixed-methods stance that lets the research question dictate method rather than paradigm (Kitchenham, Mendes, & Travassos, 2007; Page et al., 2021).
The model answers the research question, what STEM knowledge, interests, and attitudes migrant learners possess and require to become interested in STEM as a higher-education option, in two complementary moves. Identification is achieved through the KIA instrument and qualitative case evidence: knowledge items approximate understanding of core scientific constructs and inquiry processes; interest items capture information habits and engagement with STEM media and spaces; attitude items probe perceived benefits/risks, trust in institutions, and images of STEM professions (Miller, 1983, 2012; CONACYT, 2017; Minciencias y Observatorio de Ciencia y Tecnología, 2015; ANII, 2014; SENACYT, 2017). This yields an empirically grounded profile of what learners possess. Specification of requirements follows from the design and implementation layers: where instruction is culturally responsive, ICT-mediated, collaborative, and explicitly scaffolded for language and socio-emotional needs, learners encounter relevance, belonging, and authentic success early; these “mastery moments” predictably improve attitudes and interests and make visible which knowledge elements must be strengthened next (Dewsbury et al., 2022; Castillo-Montoya et al., 2023; Kim et al., 2023; Galvis & Pedraza, 2013; Palmer, 2012; Galvis Panqueva, 2021). The theory of change is therefore cyclical: inclusive design, early mastery and recognition, more positive perceptions (KIA), stronger computational-thinking practices, and higher willingness to pursue STEM studies; measurement closes the loop by indicating where to adjust tasks and scaffolds (Durant et al., 1989, 1998; Miller, 2012).
Synthesis across the layers indicates that migrant students typically begin with uneven disciplinary knowledge, situational interests clustered around technology and tangible tasks, and ambivalent attitudes shaped by perceived difficulty and limited trust; under the model, these profiles shift when teachers enact DEI-aligned routines, and students experience structured, ICT-mediated problem solving tied to their contexts. In practical terms, the model produces three audience-specific recommendations that follow from the evidence. For STEM-education researchers, make visible and theorize migrants’ situated knowledge and products emerging from training, videos, programs, prototypes, and reflections, so that scholarship actively counters anti-migration narratives and documents learning assets, not only deficits (Pain, 2004; Abrego, 2014). For primary, secondary, and middle-school teachers, guarantee continuity in learners’ formal trajectories by designing inclusive classrooms that leverage informal experiences and funds of knowledge gathered along migration routes; enact active, collaborative, language-aware routines and use ICT to personalize access and practice (Dewsbury et al., 2022; Galvis & Pedraza, 2013). For policymakers, shape migration policies that recognize migration experiences and prioritize regularization for children and families; irregular status undermines well-being and discourages sustained participation in formal schooling, blunting the motivational gains that inclusive STEM instruction can create (Abrego, 2014). By aligning equity-centered design, teacher development, learner experience, and rigorous measurement, the model not only describes what migrant learners currently know, want, and believe about STEM but also specifies and tests the instructional and policy conditions they require to translate that profile into computational thinking, strengthened interest, and a credible route into STEM higher education.Colfuturo y Ministerio de Ciencias de ColombiaDoctorad
Un giro no antropocéntrico: ¿confianza o dependencia en la IA en la investigación científica?
No full textEsta tesis examina la posibilidad de atribuir una noción robusta de confianza a los sistemas de Inteligencia Artificial (IA) utilizados en la práctica científica contemporánea. Teniendo en cuenta la opacidad epistémica esencial y las nociones tradicionales de confianza en la filosofía de la ciencia, se argumenta que dichas nociones no se pueden ni se deben trasladar a los sistemas de inteligencia artificial puesto que son entendidas como relaciones cargadas moralmente entre agentes y estos sistemas no pueden suplir dicha condición. En lugar de ello, se propone reemplazar el concepto de confianza por el de dependencia epistémica justificada, lo que permite incluir responsablemente a la inteligencia artificial en la práctica científica sin recurrir a analogías antropocéntricas que no son sostenibles.Maestrí
Conversational model to support security queries from android developers
No full textThe rapid and continuous expansion of Android as an operative system and the applications that it brings with it, has introduced constant challenges to the users security and privacy. These challenges include the existence of countless types of vulnerabilities with multiple potential some of which can only be avoided by the developers direct action. However, developers must face challenges to develop secure applications due to a lack of resources and expertise, particularly among smaller organizations, leading to risks in the security of applications. To address this problem, this paper proposes developing a conversational model based on an architecture that integrates the Retrieval-Augmented Generation (RAG) and the HyDE (Hypothetical Document Embeddings) techniques. This architecture leverages trusted knowl edge stored in an offline database that was built including exclusively academic papers and official Android developer information 1. Later, to answer a given query the information is retrieved based on semantic similarity searches. This allows the model to retrieve relevant context and generate answers that are accurate and trustable, reducing the risk of hallucina tions.
The proposed solution was evaluated using both automated metrics, such as BERT and ROUGE-L, and manual evaluation made by expert. The model showed accurate results to Android application security queries, but the information also tends to be less complete that the generated by other models, because it does not include general aspects that may be useful.
The results suggest that there is potential in combining advanced retrieval and embed ding techniques with generative models to provide accurate results, while also display the limitations that these systems have because they rely on limited information.
Still there are ways to try and mitigate these limitations, future works to do so may include approaching new sources of information that went unnoticed for this project, or focusing on a more specific topic with enough information about it to allow building a more solid modelMaestrí
Metrópolis: re-definiendo el borde.
No full textLa intervención en Metrópolis trasciende el ámbito del diseño arquitectónico tradicional para posicionarse como una herramienta de investigación urbana. A través de la eliminación de la reja, la reorganización del espacio y la jerarquización de los usos, se propone explorar cómo la transformación de un borde rígido en uno activo puede redefinir las dinámicas sociales, económicas y culturales de un entorno urbano consolidado.Pregrad
Impact of palm-oil rejuvenators on adhesion quality and moisture damage resistance of asphalt-aggregate systems
No full textMoisture damage in asphalt mixtures impacts the durability of asphalt pavements, leading to increased transportation costs and reduced service life. While anti-stripping additives are commonly used to mitigate this phenomenon, ensuring durability remains a challenge, particularly in mixtures that include new components, such as reclaimed asphalt pavement (RAP). Although rejuvenators have been proposed to restore some fundamental and performance properties of high-RAP-content mixtures, their impact on adhesion durability remains unexplored. This study assesses the effect of two palm-oil rejuvenators on the adhesion durability of asphalt-aggregate systems. Aggregate-binder specimens fabricated with one type of aggregate and multiple binders (virgin asphalt, blends of asphalt and aged RAP-binder, and rejuvenated asphalts blends) were tested using a recently developed adhesion pull-off method under dry conditions and after 3 and 7 days of water conditioning. The results demonstrate that the selected palm-oil rejuvenators effectively restore adhesion durability in asphalt-aggregate systems, highlighting their potential for enhancing not only the performance properties of the mixture, but also its durability.MaestríaMateriales asfálticosDaño por humeda
A multivariate approach to dimension estimation on manifolds with triangle area U-statistics
No full textFor data on a manifold M ⊆ ℝᵐ and a point p ∈ M, this thesis introduces a multivariate estimator that leverages angle- and triangle area-based U-statistics (U₁, V₁, V₂) to assess the intrinsic dimension of M at p. By considering the variance of angles formed by pairs of nearby data points, and the mean and variance of triangle areas formed by triplets, the proposed methodology captures the manifold's local geometry. A multivariate approach using the Mahalanobis distance ensures robust dimension estimation by incorporating covariance structure. The estimator is evaluated through testing on both simulated manifolds and real-world datasets. Robust statistical tools, such as the Minimum Volume Ellipsoid (MVE), are employed to enhance reliability.Pregrad
Derechas emergentes en Colombia: configuración y posicionamiento de la opinión en Twitter (X)
No full textEl texto analiza dos transiciones ocurridas en 2022: la transformación de Twitter en X tras la llegada de Elon Musk y el ascenso del primer gobierno de izquierda en Colombia, luego un largo periodo de hegemonía de derecha liderada en los últimos años por el uribismo. Ambas situaciones, aunque distintas, están interconectadas, ya que Twitter ha sido un espacio clave para la política colombiana. Se generan incertidumbres tanto por los cambios en la plataforma como por el giro político en el país. El estudio se enfoca en cómo Twitter ha influido en la reconfiguración de las derechas colombianas tras la llegada de Gustavo Petro y Francia Márquez al poder. Examina cómo la plataforma facilita la difusión y posicionamiento de discursos de derecha, tomando en cuenta las modificaciones impulsadas por Musk y el impacto de la comunicación digital en la política. La investigación se divide en dos partes: primero, se analiza la configuración de Twitter y su impacto en la propagación de discursos políticos; luego, se estudia la evolución de las derechas emergentes en Colombia, a través del análisis de figuras como María Fernanda Cabal y Miguel Polo Polo. Se aborda la relación entre redes sociales, comunicación política y subjetividades políticas, dentro del contexto del auge de nuevas derechas a nivel global.Maestrí
Identificación de nuevos inhibidores selectivos de la enzima AChE de la Drosophila melanogaster (mosca de la fruta) mediante métodos computacionales
No full textEl presente trabajo de investigación tuvo como objetivo principal identificar inhibidores selectivos de la enzima acetilcolinesterasa (AChE) de Drosophila melanogaster, buscando compuestos con mayor afinidad hacia esta variante enzimática en comparación con la AChE humana, como una estrategia para desarrollar alternativas menos tóxicas a los insecticidas comerciales convencionales. En el estudio se examinaron a cuatro compuestos prometedores provenientes de un cribaje de compuestos orgánicos volátiles con propiedades química similares a la tacrina, utilizando simulaciones computacionales avanzadas que incluyeron técnicas de docking, dinámica molecular (DM) y cálculos de energía libre (FEP) para evaluar las interacciones de estos compuestos con las AChE de moscas de la fruta, humanos y anguilas eléctricas. Los resultados más destacados revelaron que los ligandos AD9 (3,5-Bis(benziloxi)tolueno) y AD10 (propil 3,4,5-trihidroxibenceno-1-carboxilato) mostraron una notable selectividad, demostrando una mayor afinidad por la AChE de la mosca de la fruta en comparación con las enzimas humanas y de anguila eléctrica. Las simulaciones computacionales permitieron identificar diferencias estructurales significativas en los sitios activos de estas enzimas, lo que explica la selectividad observada. Los cálculos de energía libre de unión respaldaron estos hallazgos, evidenciando que las interacciones específicas entre los ligandos y los residuos aminoácidos del sitio activo son fundamentales para determinar la afinidad y selectividad de los compuestos. El estudio concluye que estos resultados abren nuevas perspectivas para el desarrollo de insecticidas más específicos y eficaces. La identificación de los ligandos AD9 y AD10, junto con el análisis detallado de las diferencias estructurales en los sitios activos enzimáticos, proporciona una base sólida para el diseño racional de nuevos inhibidores. Las simulaciones computacionales demostraron ser una herramienta invaluable para comprender los mecanismos de interacción molecular y predecir la actividad de nuevos compuestos. Además, los resultados sugieren que modificaciones estructurales específicas en los ligandos, especialmente en el caso del AD10, podrían mejorar aún más su afinidad y selectividad.The present work primarily aimed to identify selective inhibitors of acetylcholinesterase (AChE) from Drosophila melanogaster, seeking compounds with greater affinity towards this enzymatic variant compared to human AChE, as a strategy to develop less toxic alternatives to conventional commercial insecticides. The study examined four promising compounds from a screening of volatile organic compounds with chemical properties similar to tacrine, using advanced computational simulations that included docking techniques, molecular dynamics, and free energy calculations (FEP) to evaluate the interactions of these compounds with AChE from fruit flies, humans, and electric eels. The most notable results revealed that ligands AD9 (3,5-Bis(benzyloxy)toluene) and AD10 (propyl 3,4,5-trihydroxybenzene-1-carboxylate) showed remarkable selectivity, demonstrating higher affinity for fruit fly acetylcholinesterase compared to human and electric eel enzymes. Computational simulations allowed the identification of significant structural differences in the active sites of these enzymes, explaining the observed selectivity. Binding free energy calculations supported these findings, demonstrating that specific interactions between ligands and active site amino acid residues are fundamental in determining the compounds’ affinity and selectivity. The study concludes that these results open new perspectives for developing more specific and effective insecticides. The identification of ligands AD9 and AD10, along with a detailed analysis of structural differences in enzymatic active sites, provides a solid foundation for the rational design of new inhibitors. Computational simulations proved to be an invaluable tool for understanding molecular interaction mechanisms and predicting the activity of new compounds. Furthermore, the results suggest that specific structural modifications in the ligands, especially in the case of AD10, could further enhance their affinity and selectivityProyecto semillaMaestríaQuímica computaciona
Encuentros
No full textEncuentros comienza con una reflexión sobre la diferencia entre "borde" y "límite". Mientras que el límite es una barrera rígida, el borde es un espacio transicional donde las realidades convergen. Se abordan conceptos clave como la conexión y el encuentro, mostrando cómo la arquitectura permite que los bordes sean zonas activas en lugar de simples fronteras. Luego, se estudia un sitio específico: Suesca, un municipio de Cundinamarca, Colombia, conocido por su geografía rocosa y su valor ecológico. Se examina la Reserva Natural El Turpial y su importancia en la conservación ambiental.
A partir de esto, se propone una intervención, donde se plantean estrategias arquitectónicas como senderos y pabellones que permiten una mejor integración entre el paisaje y el ser humano. Se analiza la materialidad, los tipos de recorridos y la relación del espacio con quienes lo habitan. Finalmente, el libro concluye con una reflexión sobre el papel de la arquitectura en la mediación entre lo natural y lo construido, subrayando la importancia de los bordes como espacios de transformación y diálogo.Pregrad