35436 research outputs found
Sort by
Corpus sense: A comprehensive tool for advanced text and discourse exploration
Get PDFCorpus Sense is a web application with a focus on content and discourse analysis designed to facilitate the exploration, analysis and visualization of linguistic corpora that incorporates some advanced functionalities not available in existing software. The tool enables users to obtain useful insights with minimal effort by combining quantitative, qualitative and AI-powered features. It is designed for small to medium-sized corpora (currently up to 2.5 million tokens), permits online corpus sharing, and offers unique functionalities, such as NLP-based keyword extraction, named entity recognition, semantic search and advanced topic modelling with LLM-generated interpretable labels. The application’s interface is simple and intuitive, in an effort to make it accessible to a wide range of user profiles. This paper provides a comprehensive overview of the application’s development, architecture and applications in corpus linguistics and discourse analysis research. This description is complemented by a discussion of the integration of novel NLP-based and AI-assisted tools with traditional corpus analysis methods.Ministerio de Ciencia, Innovación y UniversidadesFunding for open access charge: Universidad de Málaga/CBU
Enhanced generation of automatically labelled image segmentation datasets by advanced style interpreter deep architectures.
Get PDFLarge image datasets with annotated pixel-level semantics are necessary to train and evaluate supervised deep-learning models. These datasets are very expensive in terms of the human effort required to build them. Still, recent developments such as DatasetGAN open the possibility of leveraging generative systems to automatically synthesise massive amounts of images along with pixel-level information. This work analyses DatasetGAN and proposes a novel architecture that utilises the semantic information of neighbouring pixels to achieve significantly better performance. Additionally, the overfitting observed in the original architecture is thoroughly investigated, and modifications are proposed to mitigate it. Furthermore, the implementation has been redesigned to greatly reduce the memory requirements of DatasetGAN, and a comprehensive study of the impact of the number of classes in the segmentation task is presented.Funding for open access charge: Universidad de Málaga / CBUAThis work is partially supported by the Ministry of Science and Innovation of Spain [grant number PID2022-136764OA-I00], project name Automated Detection of Non Lesional Focal Epilepsy by Probabilistic Diffusion Deep Neural Models. It is also partially supported by the Autonomous Government of Andalusia (Spain) under project UMA18-FEDERJA-084, project name Detection of anomalous behaviour agents by deep learning in low-cost video surveillance intelligent systems. All of them include funds from the European Regional Development Fund (ERDF)
Kit de lectura dualizada en torno a los Objetivos de Desarrollo Sostenible
Get PDFEste kit educativo surge como parte de una propuesta innovadora
desarrollada en el marco de una experiencia didáctica basada en el
Diseño Universal para el Aprendizaje (DUA), los Objetivos de Desarrollo
Sostenible (ODS) y las posibilidades que ofrece la Inteligencia Artificial (IA)
para personalizar el aprendizaje. El objetivo principal es acercar al
alumnado de Educación Primaria a problemáticas reales de nuestro
entorno a través de historias de no ficción que fomentan la conciencia
social, la comprensión lectora y la inclusión educativa.
Este kit contiene una colección de relatos breves vinculados a distintos
ODS (como el fin de la pobreza, hambre cero, salud y bienestar, igualdad
de género, entre otros), los cuales han sido adaptados a diferentes niveles
de dificultad para garantizar la accesibilidad a todo el estudiantado,
independientemente de su perfil lector. Gracias al uso de herramientas
basadas en IA, ha sido posible crear versiones múltiples de una misma
historia, respetando su esencia pero adecuándola en vocabulario,
estructura y apoyos visuales
Data for comparative study between Traditional RSVP_BCI Spellers and T9-Keyboard RSVP_BCI Spellers
No full textData were collected using BCI2000 softwareCollected data for the article: " A Rapid Serial Visual Presentation-BCI Speller Based on a T9 Keyboard: Preliminary Tests in Patients With Amyotrophic Lateral Sclerosis
Dissecting Economic Op-Eds: An Annotation Schema for Editorials in Quality Newspapers
Get PDFThis study examines the rhetorical structure and sentiment dynamics of 82 economic opinion editorials authored by prominent commentators. Using a novel annotation schema, we analyze Functional Discourse Units (FDUs), polarity, entities, and opinion holders to uncover the linguistic and rhetorical strategies employed in economic op-eds. The findings reveal a predominance of negative evaluations, highlighting the critical tone characteristic of the genre. Description/Analysis and Context FDUs are the most frequent, reflecting the op-eds’ focus on situating economic issues within broader narratives. Annotation challenges, such as gradation of sentiment and irony, emphasize the complexity of evaluating economic discourse. Despite these challenges, high inter-annotator agreement in key categories supports the robustness of the schema. This study bridges computational linguistics and discourse analysis, offering valuable insights into economic journalism while laying a foundation for future research on evaluative language and sentiment analysis in specialized corpora.Funding for open access charge: Universidad de Málaga / CBU
Involuntad
Get PDFInvoluntad trata de manifestar y hacer visible, a través de la materia, la presencia de la naturaleza en el ser humano y la manera en que este la habita con frecuencia en situaciones cotidianas y familiares para cualquier espectador, como el simple hecho de estar acostado o acurrucado. En esos momentos, de manera inconsciente, buscamos nuestro propio «núcleo», nuestro centro, una parte profunda de nuestro ser. Esta búsqueda del centro corporal al adoptar la posición fetal remite inevitablemente a las espirales geométricamente perfectas presentes en la naturaleza.
El uso de la naturaleza como material en forma de semilla, convirtiéndola en coautora de la obra, es un aspecto fundamental. La naturaleza, en su esen- cia, es lo incontrolable y no someterla buscando unos resultados específicos se convierte en un pacto en el que la obra se enriquece con un elemento natural y nacido del azar.
Al final, la posición fetal es una representación física de la vulnerabilidad y de una introspección pura a través del cuerpo, una postura que compar- timos todos los seres humanos
Propagador electromagnético 3D para antenas ópticas integradas.
Get PDFAccurate modeling of electromagnetic propagation is crucial for the design and
optimization of integrated optical antennas, with applications in chip-to-fiber coupling,
beam shaping, and free-space optical (FSO) communications. This work presents a 3D
electromagnetic propagator based on the angular spectrum method, capable of computing
the field evolution over both near-field and far-field regimes without paraxial
approximations.
The proposed approach enables precise long-distance propagation, maintaining a
integral overlap with the analytical solution for Gaussian beams over distances up to tens
of kilometers. Additionally, we provide simple analytical expressions to configure
simulation parameters, ensuring numerical accuracy and computational efficiency. The tool
is validated through two case studies: (i) long-distance Gaussian beam propagation in free
space and (ii) near-field characterization of an integrated optical antenna for fiber coupling.
The results confirm the versatility and accuracy of the proposed propagator for photonic
applications
Corpus annotation of functional discourse units for aspect‑based sentiment analysis
Get PDFAspect-based sentiment analysis (ABSA) aims to identify the sentiment associated
with specifc aspects or entities in a text. In order to facilitate the development and
evaluation of ABSA systems, it is crucial to have annotated datasets that contain
information about the aspects, entities, and the sentiments expressed towards them.
However, the amount of information in existing datasets (for example those used in
the SemEval shared tasks) is very limited. We innovate on existing corpora by introducing a multi-layered annotation schema that includes not only entities and aspects,
but also lexical items and, crucially, functional discourse units (FDUs). These FDUs
are text segments (typically sentences or clauses) that play a specifc role or function
within the overall text, such as “description”, “evaluation”, or “advice”, a type of
information which we believe can be of great help in ABSA. Our corpus focuses on
user reviews of tourist attractions (specifcally monuments) in the region of Andalusia (Spain), but the same schema can be used to annotate reviews of other domains
simply by adapting the aspects layer, which is domain-dependent. The annotation
schema is described, and the validation process is carried out on a sample of 400
reviews from this domain. Results show a substantial level of agreement among the
annotators, indicating that the schema is reliable and consistent. We go on to illustrate and discuss some difcult cases where annotation showed discrepancy among
annotators. The annotation of FDUs in the corpus is a signifcant advancement for
aspect-based sentiment analysis.Funding for open access charge: Universidad de Málaga / CBU
Analysis of On-Board Computers in CubeSats for Communication Payloads.
Get PDFOver the past few years, the number of space-related applications has increased as the acceptance of higher risk thresholds in the space industry has facilitated the widespread integration of commercial off-the-shelf (COTS) components, leading to a transformation in the design of new space missions. One of the most attractive areas of current space research is the design of miniaturized satellites, known as Cubesats, which are tempting because of their numerous applications and low design and deployment cost. Cubesats are now widely used in numerous commercial, scientific and research applications as they offer a number of interesting features in terms of development and versatility. Cubesats also provide a complementary connectivity solution to existing networks and on-board computation plays a crucial role in the design of Cubesat missions, as requirements are needed to be able to provide communications payloads that meet the objectives. This article analyzes the types of on-board computers (OBC) and key features offered by Cubesat missions in projects of all kinds aimed at providing communications via this type of satellites. Finally, it identifies future research lines for Cubesat communications in relation to OBCs
Mascarillas faciales y endoftalmitis tras inyecciones intravítreas en el período covid
Get PDFINTRODUCCIÓN: El uso extendido de las mascarillas en toda la población a raíz de la pandemia por covid supuso un cambio en la práctica clínica diaria. En el campo de la Oftalmología, concretamente al realizar inyecciones intravítreas, las mascarillas podrían aumentar el riesgo de endoftalmitis al acumular microorganismos cerca del globo ocular. Además, al redirigir el flujo exhalado hacia el ojo, el perfil de las bacterias podría variar abundando patógenos más virulentos de la orofaringe.
MATERIAL Y MÉTODOS: Se recogieron muestras de las mascarillas quirúrgicas de la zona periocular y la zona bucal de mascarillas de 40 pacientes sometidos a intravítreas. En 20 de ellas se aplicó povidona yodada sobre la zona periocular de las mascarillas faciales y en las otras 20 no se aplicó ningún tipo de tratamiento. Se cultivaron las muestras para cuantificar las UFC (Unidades Formadoras de Colonia) y se identificaron las cepas bacterianas.
RESULTADOS: La carga bacteriana en la zona periocular fue mayor en el grupo de más de 4 horas de uso en comparación con el grupo de menos de 4 horas (13,2 frente a 48,75 UFC/μL; P = 0,03). El perfil bacteriano periocular no coincidió con el recogido de la zona bucal. El uso de povidona yodada se correlacionó fuertemente con un cultivo periocular negativo (OR = 9,0, P = 0,00. IC 1,63-49,44).
CONCLUSIÓN: La carga bacteriana periocular encontrada en las mascarillas es mayor con un uso mayor a 4 horas. El origen de estas bacterias parece provenir de la propia flora periocular y no de la orofaringe. Aplicando povidona yodada a la mascarilla la carga bacteriana disminuye. El uso de mascarillas podría ser un factor de riesgo para las endoftalmitis. Por ello, para la inyección intravítrea de medicamentos es recomendable un uso menor a 4 horas y la aplicación de povidona yodada periocular