Poli[Papers] (Universitat Politècnica de València)
Not a member yet
10801 research outputs found
Sort by
Exploring AI as a Tool to Improve the English Pronunciation of Spanish-Catalan Teenagers
This study investigates how artificial intelligence (AI) can be used to improve the pronunciation skills of English as a foreign language (EFL) learners, with a specific focus on American English vowel sounds. After a brief introduction to AI concepts, the research discusses the importance of teaching and learning pronunciation, the complexities of the English vowel system, and the specific difficulties encountered by Spanish-Catalan speakers. Furthermore, the possible use of AI in second language acquisition (SLA) and in the language classroom is also addressed. To assess the efficacy of AI tools, Siri and the Speakometer app were tested with 24 Spanish-Catalan teenagers, divided into a control group using traditional methods and an experimental group using the Speakometer app. Both groups received identical explicit instruction on English vowel sounds over four sessions. All participants interacted with Siri before and after the intervention to assess AI’s speech perception while being recorded for later analysis of the participants’ speech production. Since there were no significant differences between the groups in perception or production results before and after the intervention, the findings imply that, for this sample, current AI applications do not significantly help learners improve their English pronunciation. Additionally, the study found that stimulus type, as well as specific vowel sounds, affects AI comprehension, suggesting that existing AI tools have not yet reached the efficiency level needed to be used in EFL pronunciation practice
Methodology for LCA Assessment of Heritage Buildings and their retrofitting works
In the current context of necessary reduction of environmental impact, retrofitting historical buildings is essential, but it must be specified the total carbon footprint, considering the original construction and the current restoration.
This research has a double objective: firstly, to develop a methodology to establish the carbon footprint of three historic Catalan Modernist buildings at the time of their construction; and secondly, to determine the carbon footprint of the restoration works of one of them.
An exhaustive analysis is carried on about the materials used, their extraction, transportation and manufacturing methods at the time. Following the methodological structure of contemporary EPDs and LCA, both aligned with the UNE-EN 15804 standard. This ensures methodological consistency between historical and modern datasets and provides a robust basis for comparison. Then, the embodied carbon of these buildings is compared with that of modern buildings construction, in order to establish which is higher. Finally, the carbon footprint of retrofitting one of them is also calculated.
The importance of this research lies in the methodology for the quantification of the total environmental impact of historic buildings, from their original construction to their restoration. The limitation of the study lies in the fact that only one of the three case studies has been retrofitted so far
Effectiveness of commercial text embedding models for multilingual, multi-class SaaS software classification: A practical study
In the rapidly evolving field of Software as a Service (SaaS), the accurate categorization of multilingual SaaS applications represents a significant challenge due to the inherent linguistic diversity and continuous growth in available software categories. This study investigates the application of commercial text embedding models, which transform textual data into numerical representations, for multilingual, large-scale, multi-class software classification tasks. We systematically compare various text embedding models integrated with classification algorithms, examining their predictive performance and transfer learning capabilities across multiple languages. Our experiments demonstrate that these embedding models exhibit substantial robustness and efficacy in both monolingual and cross-lingual classification contexts. Notably, a multi-layer perceptron classifier trained on bilingual datasets (French and English) using OpenAI’s text-embedding-3-large embedding model achieved high accuracy (0.90) and F1-score (0.78), even when evaluated on languages not represented in the training corpus. This research not only offers valuable insights for professionals and practitioners in the SaaS sector but also lays the groundwork for further research in advanced applications, crucial for handling the extensive textual data in the contemporary digital marketplace
Une interlocution critique : Le Corbusier dans la formation urbanistique de Louis I. Kahn
Historiography on Louis I. Kahn has largely privileged the study of his architectural production after 1950, relegating his formative years to a critical shadow. Yet this early period was decisive in shaping his urban thought, forged at the intersection between his Beaux-Arts education and the self-training of the 1930s, aimed at a disciplinary modernity centered on the city and collective housing. Within this process, Le Corbusier emerged as a central reference, providing conceptual foundations for rethinking the modern condition of the American city. Although decisive in the genesis of Kahnian urbanism, this influence has been examined only fragmentarily, without consideration of its critical depth or enduring resonances. Hence the relevance of delving into its early manifestations, mediated by the dissemination in the United States of Le Corbusier’s key theoretical texts — Vers une architecture and Urbanisme — discussed in forums such as the T-Square Club Journal and the Architectural Research Group, and crystallized in the Housing in the Rational City Plan exhibited at MoMA in 1939. In that project, Kahn synthesized and reformulated Corbusian principles, shifting the relationship between master and disciple toward a sustained intellectual interlocution that inaugurated his own projective agenda.La historiografía sobre Louis I. Kahn ha privilegiado el estudio de su producción arquitectónica posterior a 1950, relegando a una penumbra crítica su etapa formativa. Este periodo fue decisivo en la gestación de su pensamiento urbanístico, forjado en la intersección entre su educación beauxartiana y la autoformación de la década de 1930, orientada hacia una modernidad disciplinar centrada en la ciudad y la vivienda colectiva. En este proceso, Le Corbusier se constituyó en referente central, aportando fundamentos para repensar la condición moderna de la ciudad estadounidense. Aunque decisiva en la génesis del urbanismo kahniano, esta influencia ha sido examinada de manera fragmentaria, sin atender a su densidad crítica ni a sus resonancias. De ahí la pertinencia de indagar en sus primeras manifestaciones, mediadas por la difusión en Estados Unidos de textos paradigmáticos de teoría corbusieriana —Vers une architecture y Urbanisme—, discutidos en foros como el T-Square Club Journal y el Architectural Research Group, y cristalizados en el Housing in the Rational City Plan presentado en el MoMA en 1939. En él, Kahn sintetizó y reformuló principios corbusierianos, desplazando la relación entre maestro y discípulo, hacia una interlocución intelectual sostenida que inauguró una agenda proyectual propia.L’historiographie de Louis I. Kahn a largement privilégié l’étude de sa production architecturale postérieure à 1950, reléguant à une ombre critique sa période formatrice. Cette phase fut pourtant décisive dans la genèse de sa pensée urbanistique, forgée à l’intersection entre son éducation beaux-artienne et l’autoformation des années 1930, orientée vers une modernité disciplinaire centrée sur la ville et l’habitat collectif. Dans ce processus, Le Corbusier s’imposa comme référence majeure, offrant les fondements conceptuels pour repenser la condition moderne de la ville américaine. Bien que déterminante dans la formation de l’urbanisme kahnien, cette influence n’a été étudiée que de manière fragmentaire, sans prendre en compte sa densité critique ni ses résonances. D’où la pertinence d’en examiner les premières manifestations, médiatisées par la diffusion aux États-Unis des textes théoriques emblématiques de Le Corbusier — Vers une architecture et Urbanisme —, débattus dans des fórums tels que le T-Square Club Journal et l’Architectural Research Group, et cristallisées dans le Housing in the Rational City Plan exposé au MoMA en 1939. Dans ce projet, Kahn synthétisa et reformula les principes corbusiens, transformant la relation entre maître et disciple en une interlocution intellectuelle soutenue qui inaugura sa propre agenda projectuelle
From evidence to insight: designing Spain’s new digital framework for public policy evaluation
While some developed countries have made steady progress towards the institutionalization of public policy evaluation, Spain's trajectory has been more variable. Law 27/2022 represents a milestone towards this goal, establishing a new State Agency for Public Policy Evaluation (hereinafter, The Agency). This legislation builds upon prior experiences with evaluation and the successful implementation of Spending Reviews by AIReF.
Building on this milestone, the article adopts a normative institutional design approach to elaborate a forward-looking scheme for The Agency. The proposal is informed by comparative experiences yet grounded in Spain’s context. Central to this design is the integration of data analytics and artificial intelligence, which requires a renewed approach to data governance, interministerial coordination, and the articulation between centralized leadership and decentralized operational capacity.
One of the article’s core contributions is the operational definition of the Evaluation Coordination Units (hereinafter, The Units), left unexplained in the law. The Units must be conceived as essential institutional nodes within a federated architecture, enabling systematic, technically robust, and policy-relevant evaluation within ministries. Additionally, the article emphasizes that Spain already holds a valuable foundation of technical infrastructure, administrative data, professional experience, and academic expertise, which should be wisely mobilized.
Success will depend on several interrelated conditions: the consolidation of The Agency’s mandate, the development of specialized and interdisciplinary talent, the creation of a secure and interoperable data infrastructure, the establishment of effective coordination mechanisms, a transparent criteria-based system for prioritizing evaluations, the ability to communicate findings clearly, and the systematic incorporation of results into decision-making
Análisis sobre el uso de UAVs para estudiar la biota antártica: aplicaciones, impactos y perspectivas a futuro. Una revisión PRISMA
Unmanned Aerial Vehicles (UAVs) have established themselves as effective tools for studying remote areas with high precision and low environmental impact, especially in regions such as Antarctica. However, knowledge gaps persist regarding their effects on the biota from this area. Therefore, the objective of this study was to conduct a systematic review of the literature on the use of UAVs in research on Antarctic biota, evaluating their applications, impacts, and future prospects. The PRISMA protocol (a standardized tool for the clear and comprehensive presentation of systematic reviews) was applied, selecting 42 articles published between 2014 and 2024 in English and Spanish. The results show an increase in the use of UAVs since 2018, mainly in studies of marine biota (72.1%), highlighting their usefulness in population censuses, health monitoring, and behavioural analysis. The use of multirotor drones (such as the DJI Phantom 3 and 4) with RGB or multispectral sensors predominated. However, only 37.2% of the studies specify both height and flight time, limiting their replicability. Although UAVs have improved data accuracy and reduced impact on wildlife, variable behavioural responses are observed across species, underscoring the need for standardized protocols. It is concluded that UAVs are valuable tools for conservation and environmental monitoring in Antarctica, but it is crucial to standardize their use, improve sensor quality, and expand their application to less-studied regions and biological groups.Los Vehículos Aéreos No Tripulados (UAVs) se han consolidado como herramientas eficaces para estudiar áreas remotas con alta precisión y bajo impacto ambiental, especialmente en zonas como la Antártida. No obstante, persisten vacíos de información sobre sus efectos en la biota de este entorno. Por ello, el objetivo de este estudio fue realizar una revisión sistemática de la literatura sobre el uso de UAVs en investigaciones sobre la biota antártica, evaluando sus aplicaciones, impactos y perspectivas a futuro. Se aplicó el protocolo PRISMA (herramienta estandarizada para la presentación clara y completa de revisiones sistemáticas), seleccionando 42 artículos publicados entre 2014 y 2024 en inglés y español. Los resultados muestran un incremento en el uso de UAVs desde 2018, principalmente en estudios de biota marina (72,1%), destacando su utilidad en censos poblacionales, monitoreo de la salud y análisis de comportamiento. Predominó el uso de los drones multirotor (como DJI Phantom 3 y 4) con sensores RGB o multiespectrales. Sin embargo, solo el 37,2% de los estudios especificaron tanto la altura como el tiempo de vuelo, limitando su replicabilidad. Aunque los UAVs han mejorado la precisión de los datos y reducido el impacto sobre la fauna, se observan respuestas conductuales variables entre especies, lo que subraya la necesidad de establecer protocolos estandarizados. Se concluye que los UAVs son herramientas valiosas para la conservación y monitoreo ambiental en la Antártida, pero es crucial estandarizar su uso, mejorar la calidad de los sensores y expandir su aplicación a regiones y grupos biológicos menos estudiados
Reconstrucción 3D histórica y difusión digital del patrimonio arquitectónico en peligro: el caso del Monasterio de Nuestra Señora del Risco (Amavida, Ávila)
This study explores an innovative methodology to valorize and digitally disseminate endangered architectural heritage, focusing on the Monasterio de Nuestra Señora del Riscoin Amavida, Ávila, Spain. The project aims to connect the historical reconstruction of the monastery at its period of maximum relevance, in the 18thcentury, with its current state. This approach uses images generated from a three-dimensional (3D) model and a 4D virtual tour. It integrates various aspects of the building, including architectural elements, site features, and potential functions. The virtual reconstruction meticulously represents decorative elements, furniture, and interior lighting to recreate the atmosphere of prayer, work, and communal life that characterized the monastic space. The project focuses on an endangered architectonic asset, under the assumption that cultural heritage serves as a testament to human history and creativity, encapsulating the essence of past civilizations. However, contemporary challenges, such as historical structures frequently becoming abandoned, pose significant threats to this heritage. The Monasterio de Nuestra Señora del Risco, with its unique architectural and historical value, exemplifies these challenges. After a long period of abandonment, from 1835 onwards, its structure was severely damaged and the building was at risk of complete loss. This study aims to address these issues through a comprehensive digital reconstruction and dissemination strategy. The applied methodology involves several phases, including bibliographic review, fieldwork, plus the use of advanced digital techniques. The bibliographic review sourced information from various archives and historical documents; fieldwork included the documentation and analysis of the remaining architectural elements, both onsite and displaced within the surrounding villages. Photogrammetry and UAV (Unmanned Aerial Vehicle) technology captured detailed images and generated a 3D model of the monastery. Specific software (SketchUp and Lumion) created a detailed model of the monastery. This model integrates data from photogrammetric surveys and historical records, ensuring accuracy and fidelity to the original structure. The 4D virtual tour allows users to explore the monastery in both its historical and current states; this provides a comprehensive understanding of its evolution over time. The digital reconstruction reveals the monastery's architectural complexity and historical significance. Key features include the unique building layout, influenced by the rugged terrain and the integration of the sacred cave into the monastic structure. To capture this, a detailed geometric survey was required; the reason is that there are differences in terms of spatial arrangement which are notable with respect to similar examples of heritage. Among these, San Jerónimo de Guisando and Aldeanueva de Santa Cruz, used as a reference for decorations and styles should be mentioned. The project also identifies and documents displaced architectural elements, contributing to a more complete understanding of the monastery’s original form. The digital reconstruction and virtual tour serve as powerful tools for heritage conservation and education. When dealing with 4Dreconstructions, it was necessary to set the exact periods for the representation. In this case, the selected period was the 18thcentury, as it corresponded to the moment of greatest relevance in the environment, when the whole complex had already been built. Hence, this project highlights the architectural richness of the monastery, and underscores its historical and cultural significance. The careful element representation, together with the identification of dislocated assets from the monastery, provides an enriched vision of the historical legacy, promoting its valorization and conservation. The project offers an immersive experience that enhances public awareness and appreciation of the monastery's historical and cultural value. This approach also demonstrates that digital techniques have great potential to support the preservation and revitalization of endangered heritage sites. The project underscores the importance of integrating digital technologies in heritage conservation, offering a detailed description of the methodology employed for the documentation and recreation of the monastery. This methodology can serve as a model for similar initiatives aimed at preserving and disseminating endangered architectural heritage.Este trabajo presenta un enfoque innovador para la valorización y difusión del patrimonio arquitectónico en peligro, tomando como caso de estudio el Monasterio de Nuestra Señora del Risco. A través de imágenes generadas a partir de un modelo tridimensional (3D) y de una visita virtual 4D, se busca conectar la reconstrucción histórica del monasterio en su momento de máximo esplendor, el siglo XVIII, con su estado actual, mostrando las transformaciones sufridas a lo largo del tiempo. La reconstrucción virtual integra diversos aspectos del edificio: elementos arquitectónicos, características del emplazamiento y posibles funcionalidades. Asimismo, se ha incorporado a la visita virtual información relativa a la actual ubicación de elementos arquitectónicos originarios del monasterio, actualmente deslocalizados. Este enfoque demuestra las capacidades interpretativas y divulgativas de las técnicas digitales que, a pesar de sus limitaciones, contribuyen significativamente a una mejor comprensión del pasado. De igual forma, se plantea como objetivo la identificación de bienes deslocalizados originarios de un bien patrimonial en peligro. Este proyecto, además de transmitir la riqueza arquitectónica del monasterio, resalta su importancia histórica y cultural. La cuidadosa representación de los elementos decorativos, mobiliario e iluminación interior permite recrear la atmósfera de oración, trabajo y vida comunitaria que caracterizó a este espacio monástico. Con ello también se ofrece una visión enriquecida del legado histórico y se promueve su valorización y conservación. El edificio estudiado presenta una singular distribución en planta, fruto de lo accidentado del terreno y de su origen ligado al culto a la roca de la cueva. Esto ha supuesto un particular reto en la reconstrucción, precisando un análisis de los restos (incluyendo adquisición de datos geométricos con fotogrametría); además, se realizó su comparación con otros ejemplos de patrimonio abulense que, si bien presentan elementos comunes, tienen en su mayoría plantas más canónicas que el caso de estudio planteado
Reseña de: Enseñar y aprender gramática en contextos multilingües
The teaching of grammar has long been a subject of ongoing debate, oscillating between rigid approaches and more open, reflective proposals that connect grammar with critical thinking and real language use. The reviewed volume aligns with the latter perspective and is centered on the EGRAMINT project, led by Carmen Rodríguez-Gonzalo and developed by the GIEL Research Group at the University of Valencia. The work offers an innovative approach to grammar instruction in multilingual contexts, moving beyond traditional models based on rote memorization, and instead advocating for a reflective, contextualized pedagogy closely tied to writing. Structured in three parts—conceptual framework, classroom implementation, and educational research—the volume bridges theory and practice through collaborative experiences involving researchers and schoolteachers from various Spanish regions. Using didactic sequences, it promotes an interlinguistic grammar approach grounded in cross-linguistic comparison, communicative function, and metalinguistic reflection. Particularly notable is the integration of grammar teaching with writing instruction and the use of the Integrated Language Curriculum (TIL), which fosters the transfer of knowledge across languages. This work makes a valuable contribution to methodological renewal, combining a strong commitment to classroom practice, research rigor, and a transformative vision of the classroom as a space for shared construction of pedagogical knowledge.La enseñanza de la gramática ha sido objeto de debate constante, entre enfoques cerrados y propuestas abiertas que la vinculan con la reflexión crítica y el uso real del lenguaje. En esta línea se sitúa el volumen reseñado, centrado en el proyecto EGRAMINT, liderado por Carmen Rodríguez-Gonzalo y desarrollado por el Grupo GIEL de la Universitat de València. La obra plantea una propuesta innovadora para enseñar gramática en contextos multilingües, superando modelos tradicionales centrados en la memorización, y apostando por una pedagogía reflexiva, contextualizada y orientada a la escritura. Estructurado en tres partes —marco conceptual, implementación en el aula e investigación didáctica—, el volumen articula teoría y práctica mediante experiencias colaborativas entre investigadores y docentes de diversas comunidades autónomas. A través de secuencias didácticas, se promueve una gramática interlingüística basada en el contraste entre lenguas, la función comunicativa y la reflexión metalingüística. Destaca la integración con la enseñanza de la escritura y el enfoque del Tratamiento Integrado de las Lenguas (TIL), que facilita la transferencia de conocimientos entre lenguas. La obra constituye una valiosa contribución a la renovación metodológica, combinando compromiso con la práctica docente, rigor investigativo y voluntad transformadora del aula como espacio de construcción compartida de conocimiento
Validación de modelos neuronales mediante la matriz de ganancias relativas
The main objective of this work is to present a procedure that complements the traditional ones, such as the analysis of the mean square error, in the validation processes of neural models of non-linear multivariable dynamic systems (MIMO). To this purpose, a comparison is made between the degree of coupling presented by these MIMO systems, in terms of their input and output variables, with respect to the level of intrinsic coupling presented by a neural model. This proposal is based on the calculation of the relative gain matrix (MGR) as a tool to make such a comparison of the neural models and real systems. Specifically, the process of calculating the MGR for a NARX neural network is detailed. In order to determine the efficiency of this proposal, different studies are carried out by comparatively assessing the relative gain λ11 of different MIMO systems of dimension 2x2. In particular, a fully decoupled MIMO system is addressed first, followed by a coupled MIMO system. In the first phase, a mathematical study is carried out. Subsequently, with the coupled MIMO system, the study is extended to a practical approach, without considering the mathematical model of the system. The results obtained endorse the initial hypotheses of this work, making it possible to strengthen the process of adjusting neural models.Este trabajo tiene como objetivo principal presentar un procedimiento que complemente a los tradicionales, como es el análisis del error cuadrático medio, en los procesos de validación de modelos neuronales de sistemas multivariables dinámicos (MIMO) no lineales. Para ello, se lleva a cabo una comparativa entre el grado de acoplamiento que presentan dichos sistemas MIMO, entre sus variables de entrada y salida, con respecto al nivel de acoplamiento intrínseco que presenta un modelo neuronal. Esta propuesta se basa en el cálculo de la matriz de ganancias relativas (MGR) como herramienta para realizar dicha comparativa entre los modelos neuronales y los sistemas reales. Específicamente, se detalla el proceso del cálculo de la MGR para una red neuronal de tipo NARX. Con objeto de determinar la eficiencia de esta propuesta, se realizan diferentes estudios valorando comparativamente la ganancia relativa λ11 que presentan diferentes sistemas MIMO de dimensión 2x2. Concretamente, se aborda por un lado el estudio de un sistema MIMO completamente desacoplado, y por otro el de uno acoplado. En una primera fase se plantea un estudio matemático. Posteriormente, con el sistema MIMO acoplado se amplía el estudio hacia un enfoque práctico, sin considerar el modelo matemático de dicho sistema. Los resultados obtenidos refrendan las hipótesis de partida de este trabajo, permitiendo robustecer el proceso de ajuste de modelos neuronales