Pasar al contenido principal

UNA APROXIMACION HOLISTICA DE SMART DATA PARA EL ANALISIS DE DATOS GUIADO POR EL CONTEXTO: APRENDIZAJE AUTOMATICO INTELIGENTE PARA EL MODELADO Y ANALISIS DEL NEGOCIO

PID2020-112540RB-C43

•
Nombre agencia financiadora Agencia Estatal de Investigación
Acrónimo agencia financiadora AEI
Programa Programa Estatal de I+D+i Orientado a los Retos de la Sociedad
Convocatoria Proyectos de I+D+i (Generación de Conocimiento y Retos Investigación)
Año convocatoria 2020
Plan Estatal 2017-2020 Plan Estatal de Investigación Científica y Técnica y de Innovación
Centro beneficiario UNIVERSIDAD DE ALICANTE

Resultados relacionados

Resultados totales (Incluyendo duplicados): 41
Encontrada(s) 1 página(s)

RUA. Repositorio Institucional de la Universidad de Alicante
Publicación Otros (Other)

Discriminative Power of the Serious Game Attention Slackline in Children and Adolescents With and Without Attention-Deficit/Hyperactivity Disorder: Validation Study

  • Ruiz-Robledillo, Nicolás
  • Lucas, Ignacio
  • Ferrer-Cascales, Rosario
  • Albaladejo-Blázquez, Natalia
  • Sanchis, Javier
  • Trujillo, Juan
  • Albaladejo-Blázquez, Natalia|||/items/27137052-c062-49fe-af8f-40a5f68c2e54
  • Ferrer-Cascales, Rosario|||/items/af59398b-f91f-4da8-8fbb-b9774ee8776f
  • Ruiz-Robledillo, Nicolás|||/items/0c5e4dff-8b8f-4a95-8239-761854dae13d
  • Sanchis, Javier|||/items/223d926e-7f4e-416b-89b9-2ee8648e6fcc
  • Trujillo, Juan|||/items/6b71a8e1-3f53-45b5-b548-767fac0d66aa
Background: Attention-deficit/hyperactivity disorder (ADHD) is a prevalent neurodevelopmental condition characterized by inattention, hyperactivity, and impulsivity, significantly impacting the psychological, social, and academic well-being of affected children and adolescents. Traditional ADHD diagnostic methods often rely on subjective reports, which can be biased. Recent advancements in serious games offer the potential for objective assessment tools. Objective: This study aimed to evaluate the discriminative power and concurrent validity of the serious game Attention Slackline in distinguishing children and adolescents with ADHD from those without the condition and in correlating game performance with standardized ADHD assessment scales. Methods: A sample of 32 children and adolescents diagnosed with ADHD and 39 healthy controls participated in the study. Participants were divided into 2 age groups: children (aged 6‐11 years) and adolescents (aged 12‐17 years). The serious game Attention Slackline was administered alongside established ADHD assessment scales, including the Child and Adolescent Assessment System and the ADHD Rating Scale IV. Group differences were analyzed using multivariate analysis of covariance, and effect sizes were reported using Cohen d. Correlations between game performance and ADHD symptoms were calculated using Pearson r. Results: Children with ADHD demonstrated significantly worse performance in Attention Slackline than the controls (t65=−2.26; P=.03; |d|=0.901), whereas no significant differences were observed in adolescents (t65=0.75; P=.73; |d|=0.191). Task performance was negatively correlated with family-reported hyperactivity/impulsivity symptoms in children across both tests (r=−0.43 and r=−0.51), but no significant correlations were observed in adolescents. Conclusions: The findings support the validity of Attention Slackline for assessing hyperactivity/impulsivity symptoms in children with ADHD. However, its efficacy decreases in adolescents, potentially due to developmental factors, such as compensatory strategies and ceiling effects in task performance. The gamified nature of the tool enhances engagement, which is crucial for young populations, while maintaining its diagnostic utility in measuring impulsivity. The age-dependent validity aligns with previous research indicating that continuous performance test paradigms are less effective in older populations due to developmental maturation. Attention Slackline shows potential as a complementary tool for ADHD diagnosis in children, offering an engaging and objective assessment of hyperactivity/impulsivity. Future research should aim to establish clinical cutoff points and refine the task’s complexity to align with individual characteristics.
RUA. Repositorio Institucional de la Universidad de Alicante
Publicación Otros (Other)

A Decision Support System Based on multi-head convolutional and Recurrent Neural Networks for assisting physicians in diagnosing ADHD

  • Sanchis, Javier
  • Teruel, Miguel A.
  • Trujillo, Juan
  • Sanchis, Javier|||/items/223d926e-7f4e-416b-89b9-2ee8648e6fcc
  • Teruel, Miguel A.|||/items/72e39554-945f-4f46-9aea-dbce9f65e43b
  • Trujillo, Juan|||/items/6b71a8e1-3f53-45b5-b548-767fac0d66aa
Background: Attention-Deficit Hyperactivity Disorder (ADHD) is highly prevalent among children and adolescents. Traditional diagnostic methods are subjective and time-consuming, underscoring the need for more objective diagnostic tools. Electroencephalography (EEG) has emerged as a promising biomarker for detecting ADHD. This study proposes MCRNet, a Multi-head Convolutional and Recurrent Neural Network, for aiding in ADHD detection using EEG and Deep Learning (DL) techniques. Method: MCRNet integrates a parallel architecture of two modules, convolutional and recurrent neural networks. The convolutional module introduces an innovative two-stage multi-head approach for enhanced feature extraction. The model was evaluated using cross-subject validation ensuring its applicability to new, unseen patients. Results: The model achieved an accuracy of 94.87% and a recall of 98.33%, indicating high reliability in identifying ADHD cases. MCRNet outperforms existing methodologies that employ raw EEG signals as input with cross-subject validation, offering an objective and reliable tool for ADHD diagnosis. Conclusions: MCRNet shows potential in reliably aiding ADHD diagnosis. Its two-stage multi-head approach enhances feature extraction and classification from raw EEG signals. Future work should focus on MCRNet’s explainability and test its efficacy on additional EEG datasets.
RUA. Repositorio Institucional de la Universidad de Alicante
Publicación Otros (Other)

Beyond TPC-DS, a benchmark for Big Data OLAP systems (BDOLAP-Bench)

  • Tardío, Roberto
  • Maté, Alejandro
  • Trujillo, Juan
  • Maté, Alejandro|||/items/93f2089c-972b-4620-bfbe-c0e3cc4a6fa3
  • Tardío, Roberto|||/items/7872c182-ff50-4097-9ef8-1e06609c666d
  • Trujillo, Juan|||/items/6b71a8e1-3f53-45b5-b548-767fac0d66aa
Online Analytical Processing (OLAP) systems with Big Data support allow storing tables of up to tens of billions of rows or terabytes of data. At the same time, these tools allow the execution of analytical queries with interactive response times, thus making them suitable for the implementation of Business Intelligence applications. However, since there can be significant differences in query and data loading performance between current Big Data OLAP tools, it is worthwhile to evaluate and compare them using a benchmark. But we identified that none of the existing approaches are really suitable for this type of system. To address this, in this research we propose a new benchmark specifically designed for Big Data OLAP systems and based on the widely adopted TPC-DS benchmark. To overcome TPC-DS inadequacy, we propose (i) a set of transformations to support the implementation of its sales data mart on any current Big Data OLAP system, (ii) a choice of 16 genuine OLAP queries, and (iii) an improved data maintenance performance metric. Moreover, we validated our benchmark through its implementation on four representative systems.
RUA. Repositorio Institucional de la Universidad de Alicante
Publicación Otros (Other)

Attention Mistakes: Juego serio para la ayuda de la detección del TDAH mediante Realidad Virtual

  • Maciá, Marina
  • Panagiotidis-Arrizabalaga, Alejandro
  • Sanchis, Javier
  • Teruel, Miguel A.
  • Trujillo, Juan
  • Maciá, Marina|||/items/5e045f47-05d6-43ac-a061-4073cad75e70
  • Panagiotidis-Arrizabalaga, Alejandro|||/items/638032a1-01c0-4559-a0b5-1b28e50d8786
  • Sanchis, Javier|||/items/223d926e-7f4e-416b-89b9-2ee8648e6fcc
  • Teruel, Miguel A.|||/items/72e39554-945f-4f46-9aea-dbce9f65e43b
  • Trujillo, Juan|||/items/6b71a8e1-3f53-45b5-b548-767fac0d66aa
Este artículo introduce el juego serio Attention Mistakes, una herramienta de apoyo para ser empleada en el diagnóstico del Trastorno por Déficit de Atención e Hiperactividad (TDAH) en adolescentes de 6 a 18 años, mediante Realidad Virtual. Esta innovadora aproximación permite recoger no solo los datos de la mecánica del juego si no también datos biométricos tales como Encefalografía, Seguimiento Ocular y Actividad Electrodermal en tiempo real. Con la finalidad de obtener un producto similar a la vida real, este juego ofrece una evaluación más precisa al simular acciones que podemos encontrar en un entorno real de trabajo, mejorando así la comprensión de las capacidades cognitivas de los evaluados. Attention Mistakes establece un avance en la detección del TDAH, al utilizar la recopilación de datos tanto de la mecánica del juego como de parámetros biométricos para implementar técnicas de inteligencia artificial y aprendizaje profundo. Este enfoque innovador facilita un diagnóstico más preciso y abre nuevas vías para la intervención efectiva en el trastorno.
RUA. Repositorio Institucional de la Universidad de Alicante
Software de investigación Software (Software)

2D Very Deep Convolutional Neural Network for Carotid Artery Calcification Detection on 3D Patient's CT-Scan

  • Reina Reina, Alejandro
  • Maté, Alejandro
  • Trujillo, Juan
  • Maté, Alejandro|||/items/93f2089c-972b-4620-bfbe-c0e3cc4a6fa3
  • Reina Reina, Alejandro|||/items/3dec779b-9915-4612-ba90-565bb21ad2a0
  • Trujillo, Juan|||/items/6b71a8e1-3f53-45b5-b548-767fac0d66aa
Modelo de red neuronal que procesa un CT-Scan realizado sobre la zona del cuello para la detección de la calcificación en la arteria carótida.
RUA. Repositorio Institucional de la Universidad de Alicante
Publicación Otros (Other)

An integrated requirements framework for analytical and AI projects

  • Trujillo, Juan
  • Lavalle, Ana
  • Reina Reina, Alejandro
  • García Carrasco, Jorge
  • Maté, Alejandro
  • Maaß, Wolfgang
  • Trujillo, Juan|||/items/6b71a8e1-3f53-45b5-b548-767fac0d66aa
  • Lavalle, Ana|||/items/bd9039d2-f3e3-4ad6-8797-8c6874286dcf
  • Reina Reina, Alejandro|||/items/3dec779b-9915-4612-ba90-565bb21ad2a0
  • Maté, Alejandro|||/items/93f2089c-972b-4620-bfbe-c0e3cc4a6fa3
  • García Carrasco, Jorge|||/items/cf588298-eb32-4b82-9b26-794828afb69c
To this day, the requirements of data warehouses, user visualizations and ML projects have been tackled in an independent manner, ignoring the possible cross-requirements, collective constraints and dependencies between the outputs of the different systems that should be taken into account to ensure a successful analytical project. In this work, we take a holistic approach and propose a methodology that supports modeling and subsequent analysis while taking into account these three aspects. This methodology has several advantages, mainly that (i) it enables us to identify possible conflicts between actors on different tasks that are overlooked if the systems are treated in an isolated manner and (ii) this holistic view enables modeling multi-company systems, where the information or even the analytical results can be provided by third-parties, identifying key participants in federated environments. After presenting the required formalism to carry out this kind of analysis, we showcase it on a real-world running example of the tourism sector.
RUA. Repositorio Institucional de la Universidad de Alicante
Publicación Otros (Other)

Sensing technologies and machine learning methods for emotion recognition in autism: Systematic review

  • Banos, Oresti
  • Comas-González, Zhoe
  • Medina, Javier
  • Polo-Rodríguez, Aurora
  • Gil, David
  • Peral, Jesús
  • Amador, Sandra
  • Villalonga, Claudia
  • Amador, Sandra|||/items/c62eecad-f696-4561-85f9-bd575e1a017c
  • Gil, David|||/items/90b814a5-7a43-4cc4-bd20-9a77d5ff0f08
  • Peral, Jesús|||/items/c944ebfe-a23a-4bde-b779-e2054a62d1a7
Background: Human Emotion Recognition (HER) has been a popular field of study in the past years. Despite the great progresses made so far, relatively little attention has been paid to the use of HER in autism. People with autism are known to face problems with daily social communication and the prototypical interpretation of emotional responses, which are most frequently exerted via facial expressions. This poses significant practical challenges to the application of regular HER systems, which are normally developed for and by neurotypical people. Objective: This study reviews the literature on the use of HER systems in autism, particularly with respect to sensing technologies and machine learning methods, as to identify existing barriers and possible future directions. Methods: We conducted a systematic review of articles published between January 2011 and June 2023 according to the 2020 PRISMA guidelines. Manuscripts were identified through searching Web of Science and Scopus databases. Manuscripts were included when related to emotion recognition, used sensors and machine learning techniques, and involved children with autism, young, or adults. Results: The search yielded 346 articles. A total of 65 publications met the eligibility criteria and were included in the review. Conclusions: Studies predominantly used facial expression techniques as the emotion recognition method. Consequently, video cameras were the most widely used devices across studies, although a growing trend in the use of physiological sensors was observed lately. Happiness, sadness, anger, fear, disgust, and surprise were most frequently addressed. Classical supervised machine learning techniques were primarily used at the expense of unsupervised approaches or more recent deep learning models. Studies focused on autism in a broad sense but limited efforts have been directed towards more specific disorders of the spectrum. Privacy or security issues were seldom addressed, and if so, at a rather insufficient level of detail.
RUA. Repositorio Institucional de la Universidad de Alicante
Publicación Otros (Other)

A benchmark of Spanish language datasets for computationally driven research

  • Candela, Gustavo
  • Sáez Fernández, María Dolores
  • Escobar Esteban, María Pilar
  • Marco Such, Manuel
  • Candela, Gustavo|||/items/df3bee46-829c-44a2-acd0-49044a591ffa
  • Escobar Esteban, María Pilar|||/items/e08f4b8d-724a-444f-93d3-56f4e07864a7
  • Marco Such, Manuel|||/items/f781969d-dee8-453a-954b-d12cd9d165f8
  • Sáez Fernández, María Dolores|||/items/36aa1ebf-ac25-45ed-bced-6888fa74f638
In the domain of Galleries, Libraries, Archives and Museums (GLAM) institutions, creative and innovative tools and methodologies for content delivery and user engagement have recently gained international attention. New methods have been proposed to publish digital collections as datasets amenable to computational use. Standardised benchmarks can be useful to broaden the scope of machine-actionable collections and to promote cultural and linguistic diversity. In this article, we propose a methodology to select datasets for computationally driven research applied to Spanish text corpora. This work seeks to encourage Spanish and Latin American institutions to publish machine-actionable collections based on best practices and avoiding common mistakes.
RUA. Repositorio Institucional de la Universidad de Alicante
Publicación Otros (Other)

Law Modeling for Fairness Requirements Elicitation in Artificial Intelligence Systems

  • Lavalle, Ana
  • Maté, Alejandro
  • Trujillo, Juan
  • García Carrasco, Jorge
  • Lavalle, Ana|||/items/bd9039d2-f3e3-4ad6-8797-8c6874286dcf
  • Maté, Alejandro|||/items/93f2089c-972b-4620-bfbe-c0e3cc4a6fa3
  • Trujillo, Juan|||/items/6b71a8e1-3f53-45b5-b548-767fac0d66aa
  • García Carrasco, Jorge|||/items/cf588298-eb32-4b82-9b26-794828afb69c
As Artificial Intelligence (AI) algorithms become widespread, concerns rise regarding their potential discrimination using protected attributes such as race, sex, or age among others. Fairness is a quality highly desired by society, however, it can be really difficult to achieve. Identifying protected attributes and trying to maintain fairness is both an important as well as a challenging task. In this paper, we propose a dynamic framework that considers the project context together with law modeling in order to identify protected attributes that threaten the fairness of AI models. This leads to a conscious evaluation of both the accuracy and fairness of the AI solutions developed. To this aim, we propose to model legal requirements by using Nòmos 3, which allows us to capture the legal requirements that should be fulfilled into legal contexts that can be loaded into our framework. By following our proposal we (i) map the duties in legal requirements to attributes in the used dataset, identifying protected attributes and providing traceability, (ii) help users in the selection of the definition of fairness that best suits the context at hand, (iii) represent the output of AI models visually in order to allow users to interpret how correct and fair are the decisions achieved by the model. To show the applicability of our proposal, we exemplify its application through a illustrative use case.
RUA. Repositorio Institucional de la Universidad de Alicante
Publicación Otros (Other)

Multiscale deep learning convolutional neural network for ADHD detection using EEG

  • Sanchis, Javier
  • Kechadi, Tahar
  • Teruel, Miguel A.
  • Trujillo, Juan
  • Sanchis, Javier|||/items/223d926e-7f4e-416b-89b9-2ee8648e6fcc
  • Teruel, Miguel A.|||/items/72e39554-945f-4f46-9aea-dbce9f65e43b
  • Trujillo, Juan|||/items/6b71a8e1-3f53-45b5-b548-767fac0d66aa
Attention-Deficit/Hyperactivity Disorder (ADHD) is a common neurodevelopmental condition in children. Electroencephalography (EEG) may support objective assessment, but subject-independent ADHD classification remains challenging because EEG varies across participants and recording segments. We developed the EEG Multiscale Convolutional Network (EEG-MSCNet), a multiscale convolutional architecture for pediatric EEG classification, and evaluated it using subject-level leave-one-subject-out (LOSO) validation on a public EEG dataset. The study also examined preprocessing choices, architectural ablations, channel-region subsets, computational complexity, and repeated-run stability. The final EEG-MSCNet evaluation achieved an Accuracy of 87.6%, F1-score of 87.6%, and area under the receiver operating characteristic curve (AUROC) of 88.2%. Artifact subspace reconstruction (ASR) combined with Fully Automated Statistical Thresholding for EEG artifact Rejection (FASTER) emerged as the strongest preprocessing configuration. The multiscale architecture showed stronger observed Accuracy and F1-score than reduced architectural variants, and the full montage provided the strongest channel configuration. These findings show that subject-level validation, preprocessing, and architectural decisions materially influence EEG-based ADHD classification. EEG-MSCNet contributes both a model and a systematic evaluation framework, supported by source code, configurations, logs, and the public dataset reference for reproducible comparison.
RUA. Repositorio Institucional de la Universidad de Alicante
Publicación Otros (Other)

Feature engineering of EEG applied to mental disorders: a systematic mapping study

  • García-Ponsoda, Sandra
  • García Carrasco, Jorge
  • Teruel, Miguel A.
  • Maté, Alejandro
  • Trujillo, Juan
  • García-Ponsoda, Sandra|||/items/b1adddd3-e77f-4ee6-bc2e-7b00acbae17c
  • Maté, Alejandro|||/items/93f2089c-972b-4620-bfbe-c0e3cc4a6fa3
  • Teruel, Miguel A.|||/items/72e39554-945f-4f46-9aea-dbce9f65e43b
  • Trujillo, Juan|||/items/6b71a8e1-3f53-45b5-b548-767fac0d66aa
  • García Carrasco, Jorge|||/items/cf588298-eb32-4b82-9b26-794828afb69c
Around a third of the total population of Europe suffers from mental disorders. The use of electroencephalography (EEG) together with Machine Learning (ML) algorithms to diagnose mental disorders has recently been shown to be a prominent research area, as exposed by several reviews focused on the field. Nevertheless, previous to the application of ML algorithms, EEG data should be correctly preprocessed and prepared via Feature Engineering (FE). In fact, the choice of FE techniques can make the difference between an unusable ML model and a simple, effective model. In other words, it can be said that FE is crucial, especially when using complex, non-stationary data such as EEG. To this aim, in this paper we present a Systematic Mapping Study (SMS) focused on FE from EEG data used to identify mental disorders. Our SMS covers more than 900 papers, making it one of the most comprehensive to date, to the best of our knowledge. We gathered the mental disorder addressed, all the FE techniques used, and the Artificial Intelligence (AI) algorithm applied for classification from each paper. Our main contributions are: (i) we offer a starting point for new researchers on these topics, (ii) we extract the most used FE techniques to classify mental disorders, (iii) we show several graphical distributions of all used techniques, and (iv) we provide critical conclusions for detecting mental disorders. To provide a better overview of existing techniques, the FE process is divided into three parts: (i) signal transformation, (ii) feature extraction, and (iii) feature selection. Moreover, we classify and analyze the distribution of existing papers according to the mental disorder they treat, the FE processes used, and the ML techniques applied. As a result, we provide a valuable reference for the scientific community to identify which techniques have been proven and tested and where the gaps are located in the current state of the art.
RUA. Repositorio Institucional de la Universidad de Alicante
Publicación Otros (Other)

Practical Methods for the Mechanistic Interpretability of Large Language Models

  • García Carrasco, Jorge
  • García Carrasco, Jorge|||/items/cf588298-eb32-4b82-9b26-794828afb69c
Los Large Language Models (LLM) han demostrado una capacidad asombrosa en una amplia gama de tareas relacionadas con el lenguaje, lo que los convierte en herramientas de gran interés. Sin embargo, debido a su elevado número de parámetros e interacciones internas, este tipo de modelos se comporta como una caja negra. Esto representa un problema importante en contextos donde una predicción incorrecta puede acarrear un coste elevado. En estos casos, resulta crucial comprender el mecanismo o proceso que subyace a una predicción para poder anticipar posibles errores. Por este motivo, el campo de investigación de la interpretabilidad ha adquirido una relevancia creciente. En particular, el reciente enfoque de la interpretabilidad mecanística ha planteado una nueva manera de abordar este problema, al centrarse en la inspección del interior de los modelos con el fin de descomponer su funcionamiento de forma comprensible. En esta línea, diversos trabajos han presentado evidencia de que, para una tarea específica, es posible localizar e interpretar el subconjunto del modelo responsable de llevarla a cabo. Esto abre un abanico de oportunidades de investigación muy prometedoras. El objetivo de la presente tesis es contribuir al emergente campo de la interpretabilidad mecanística, con especial atención a las aplicaciones prácticas que derivan de estos avances. En concreto, la tesis se centra en: (i) extender los métodos actuales de interpretabilidad mecanística para abordar tareas más complejas, en las que la predicción se compone de múltiples tokens; (ii) desarrollar un método para localizar vulnerabilidades dentro de los LLM; y (iii) diseñar un procedimiento que permita extraer el subconjunto del LLM responsable de una tarea concreta, con el fin de utilizarlo de manera independiente. Los resultados obtenidos durante la realización de esta tesis doctoral se han plasmado en una serie de artículos científicos presentados en congresos internacionales y publicados en revistas de alto impacto. Por este motivo, la tesis se presenta en formato de compendio de publicaciones.
RUA. Repositorio Institucional de la Universidad de Alicante
Publicación Otros (Other)

A Methodology based on Rebalancing Techniques to measure and improve Fairness in Artificial Intelligence algorithms

  • Lavalle, Ana
  • Maté, Alejandro
  • Trujillo, Juan
  • García Carrasco, Jorge
  • Lavalle, Ana|||/items/bd9039d2-f3e3-4ad6-8797-8c6874286dcf
  • Maté, Alejandro|||/items/93f2089c-972b-4620-bfbe-c0e3cc4a6fa3
  • Trujillo, Juan|||/items/6b71a8e1-3f53-45b5-b548-767fac0d66aa
  • García Carrasco, Jorge|||/items/cf588298-eb32-4b82-9b26-794828afb69c
Artificial Intelligence (AI) has become one of the key drivers for the next decade. As important decisions are increasingly supported or directly made by AI systems, concerns regarding the rationale and fairness in their outputs are becoming more and more prominent nowadays. Following the recent interest in fairer predictions, several metrics for measuring fairness have been proposed, leading to different objectives which may need to be addressed in different fashion. In this paper, we propose (i) a methodology for analyzing and improving fairness in AI predictions by selecting sensitive attributes that should be protected; (ii) We analyze how the most common rebalance approaches affect the fairness of AI predictions and how they compare to the alternatives of removing or creating separate classifiers for each group within a protected attribute. Finally, (iii) our methodology generates a set of tables that can be easily computed for choosing the best alternative in each particular case. The main advantage of our methodology is that it allows AI practitioners to measure and improve fairness in AI algorithms in a systematic way. In order to check our proposal, we have properly applied it to the COMPAS dataset, which has been widely demonstrated to be biased by several previous studies.
RUA. Repositorio Institucional de la Universidad de Alicante
Publicación Otros (Other)

A novel approach to identify the brain regions that best classify ADHD by means of EEG and deep learning

  • Sanchis, Javier
  • García-Ponsoda, Sandra
  • Teruel, Miguel A.
  • Trujillo, Juan
  • Song, Il-Yeol
  • García-Ponsoda, Sandra|||/items/b1adddd3-e77f-4ee6-bc2e-7b00acbae17c
  • Sanchis, Javier|||/items/223d926e-7f4e-416b-89b9-2ee8648e6fcc
  • Teruel, Miguel A.|||/items/72e39554-945f-4f46-9aea-dbce9f65e43b
  • Trujillo, Juan|||/items/6b71a8e1-3f53-45b5-b548-767fac0d66aa
Objective: Attention-Deficit Hyperactivity Disorder (ADHD) is one of the most widespread neurodevelopmental disorders diagnosed in childhood. ADHD is diagnosed by following the guidelines of Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5). According to DSM-5, ADHD has not yet identified a specific cause, and thus researchers continue to investigate this field. Therefore, the primary objective of this work is to present a study to find the subset of channels or brain regions that best classify ADHD vs Typically Developing children by means of Electroencephalograms (EEG). Methods: To achieve this goal, we present a novel approach to identify the brain regions that best classify ADHD using EEG and Deep Learning (DL). First, we perform a filtering and artefact removal process on the EEG signal. Then we generate different subsets of EEG channels depending on their location on the scalp (hemispheres, lobes, sets of lobes and single channels) and using backward and forward stepwise feature selection methods. Finally, we feed the DL neural network with each set, and compute the f1-score. Results and Conclusions: Based on the obtained results, the Frontal Lobe (FL) (0.8081 f1-score) and the Left Hemisphere (LH) (0.8056 f1-score) provide more significant information detecting individuals with ADHD, than using the entire set of EEG Channels (0.8067 f1-score). However, when combining the Temporal, Parietal and Occipital Lobes (TL, PL, OL), better results (0.8097 f1-score) were obtained compared with using only the FL and LH subsets. The best performance was obtained using Feature Selection Methods. In the case of the Backward Stepwise Feature Selection method, a combination of 14 EEG channels yielded a 0.8281 f1-score. Similarly, using the Forward Stepwise Feature Selection method, a combination of 11 EEG channels yielded a 0.8271 f1-score. These findings hold significant value for physicians in the quest to better understand the underlying causes of ADHD.
RUA. Repositorio Institucional de la Universidad de Alicante
Publicación Otros (Other)

Tracking daily paths in home contexts with RSSI fingerprinting based on UWB through deep learning models

  • Polo-Rodríguez, Aurora
  • Valera, Juan Carlos
  • Peral, Jesús
  • Gil, David
  • Medina, Javier
  • Gil, David|||/items/90b814a5-7a43-4cc4-bd20-9a77d5ff0f08
  • Peral, Jesús|||/items/c944ebfe-a23a-4bde-b779-e2054a62d1a7
The field of human activity recognition has evolved significantly, driven largely by advancements in Internet of Things (IoT) device technology, particularly in personal devices. This study investigates the use of ultra-wideband (UWB) technology for tracking inhabitant paths in home environments using deep learning models. UWB technology estimates user locations via time-of-flight and time-difference-of-arrival methods, which are significantly affected by the presence of walls and obstacles in real environments, reducing their precision. To address these challenges, we propose a fingerprinting-based approach utilizing received signal strength indicator (RSSI) data collected from inhabitants in two flats (60 m and 100 m) while performing daily activities. We compare the performance of convolutional neural network (CNN), long short-term memory (LSTM), and hybrid CNN+LSTM models, as well as the use of Bluetooth technology. Additionally, we evaluate the impact of the type and duration of the temporal window (future, past, or a combination of both). Our results demonstrate a mean absolute error close to 50 cm, highlighting the superiority of the hybrid model in providing accurate location estimates, thus facilitating its application in daily human activity recognition in residential settings.
RUA. Repositorio Institucional de la Universidad de Alicante
Publicación Otros (Other)

Measuring attention of ADHD patients by means of a computer game featuring biometrical data gathering

  • Teruel, Miguel A.
  • Sanchis, Javier
  • Ruiz-Robledillo, Nicolás
  • Albaladejo-Blázquez, Natalia
  • Ferrer-Cascales, Rosario
  • Trujillo, Juan
  • Albaladejo-Blázquez, Natalia|||/items/27137052-c062-49fe-af8f-40a5f68c2e54
  • Ferrer-Cascales, Rosario|||/items/af59398b-f91f-4da8-8fbb-b9774ee8776f
  • Ruiz-Robledillo, Nicolás|||/items/0c5e4dff-8b8f-4a95-8239-761854dae13d
  • Sanchis, Javier|||/items/223d926e-7f4e-416b-89b9-2ee8648e6fcc
  • Teruel, Miguel A.|||/items/72e39554-945f-4f46-9aea-dbce9f65e43b
  • Trujillo, Juan|||/items/6b71a8e1-3f53-45b5-b548-767fac0d66aa
ADHD is a neurodevelopmental disorder diagnosed mainly in children, marked by inattention and hyperactivity-impulsivity. The symptoms are highly variable, such as different ages of onset and potential comorbidities, contributing to frequent misdiagnoses. Professionals note a gap in modern diagnostic tools, making accurate identification challenging. To address this, recent studies recommend gamification for better ADHD diagnosis and treatment, though further research is essential to confirm its efficacy. This work aims to create a serious game, namely “Attention Slackline", to assess attention levels. The game, designed with expert input, requires players to concentrate on a specific point to recognize specific patterns while managing distractions. A controlled experiment tested its precision, and results were compared with established attention tests by a correlation analysis. Statistical analysis confirmed the game's validity, especially in tracking attention through correct responses and errors. Preliminary evidence suggests that “Attention Slackline" may serve as a credible instrument for the assessment of attentional capacities in individuals with ADHD, given that its outcomes have been empirically shown to correlate with those derived from a well-established attention assessment methodology.
RUA. Repositorio Institucional de la Universidad de Alicante
Publicación Otros (Other)

Una propuesta para la captura e implementación de requisitos para eXplainable Artificial Intelligence alineada con la Arquitectura Dirigida por Modelos

  • Navarro, Álvaro
  • Navarro, Álvaro|||/items/7ee415a4-651b-46a5-aeb9-32f9c9931cf8
La era de los datos ha dado pie a que la Inteligencia Artificial (IA) avance a una velocidad sin parangón tanto a nivel científico como, al mismo tiempo, su impacto en la sociedad y la ciudadanía que compone ésta. Pero, desgraciadamente, la aplicación de sistemas de IA no siempre obtiene los resultados esperados. Discriminaciones o errores no previstos son solamente algunas de dichas consecuencias negativas. Para poder solucionar dichos problemas -y mejorar así la vida de la ciudadanía- se necesita que los sistemas de IA sean interpretables y así conseguir que (i) los desarrolladores o expertos en IA puedan verificar lo que ocurre de una manera más completa -observando qué ocurre dentro del sistema y no únicamente su salida y la evaluación de ésta- antes de su aplicación real, y (ii) la ciudadanía que usa dichos sistemas pueda confiar en éstos. No obstante, muchos de los modelos de Aprendizaje Automático y aprendizaje profundo (ML y DL por sus siglas en inglés Machine Learning y Deep Learning, respectivamente), que son subáreas de la IA, presentan una falta de transparencia respecto a sus procesos de toma de decisiones, es decir, el rápido avance de la IA ha conllevado que los modelos -cada vez más complejos- presenten un carácter opaco o de caja negra que limita el use de éstos. Esto es particularmente grave cuando se observa que la mayoría de los usuarios en la sociedad no tienen experiencia en estas áreas pese a estar sujetos en su día a día a los resultados de los modelos de IA. Para poder solucionar este problema surgió el campo de la explicabilidad de la IA (XAI, por eXplainable Artificial Intelligence}). De esta forma, gracias a la aplicación de XAI es posible generar explicaciones -que se incluirán en interfaces explicables entendibles- y así se satisfará con información de cómo el sistema de IA ha tomado sus decisiones. Sin embargo, es un campo aún no explorado lo suficiente y que presenta varias carencias. En primer lugar, existe una carencia de conceptualizaciones y formalismos para ayudar al análisis y selección de técnicas de XAI, lo que imprescindible debido a: por un lado, por la complejidad de dicha tarea debido a todas las técnicas existentes y los aspectos que afectan a cuáles son -más- apropiadas; por otro lado, la falta de desarrolladores o expertos en XAI conlleva que, generalmente, las tareas XAI sean ejecutadas por desarrolladores o expertos en IA, quienes -a menudo- carecen de un conocimiento exhaustivo del área de la explicabilidad. Además, incluso en los escenarios donde XAI se aplica con resultados positivos, normalmente no se realiza una evaluación teniendo en cuenta (i) la fiabilidad y precisión del sistema en su conjunto, y (ii) los destinatarios objetivo de la explicabilidad y su nivel de entendimiento. Es decir, a menudo se aplica XAI para poder verificar que el sistema de IA funciona correctamente (en concreto, esto ocurre en los casos en los que el objetivo de XAI es de desarrollo y no en otras cuestiones como las del dominio), pero si no se evalúa la técnica de explicabilidad aplicada, no se resuelve el problema, sino que se traslada a un plano distinto. Debido a ello, pese a que los formalismos matemáticos ligados -en ocasiones- a la técnica XAI en cuestión y su explicación generada pueden satisfacer a los desarrolladores o expertos XAI en algunos escenarios, (i) la inevitable implicación de usuarios no expertos en IA y XAI, y (ii) la falta de criterios para comparar unas explicaciones con otras -y, de esta forma, las técnicas que las han generado- motivan la necesidad de un consenso sobre cómo evaluar la explicabilidad. De esta forma, diseñar dicho consenso es uno de los principales objetivos del campo de XAI. Además, éste debe ser aplicable en los distintos escenarios de IA donde XAI se incorpora con independencia del caso de estudio, y cuantificable de manera holística para así permitir la comparación entre las distintas explicaciones y técnicas. Motivado por ello, en la presente tesis doctoral hemos incluido como núcleo la investigación realizada para satisfacer distintos objetivos relevantes en el área actual de XAI. En concreto, afrontamos el reto de: (i) facilitar la captura y análisis de los usuarios implicados en XAI junto con sus objetivos y necesidades; (ii) facilitar la selección de las técnicas XAI más apropiadas; (iii) desarrollar un enfoque para el diseño y la implementación de XAI que tenga en cuenta quienes son los usuarios finales directamente desde sus requisitos (es decir, derivar de (i) a (ii)); (iv) analizar y evaluar la justicia y los sesgos de la IA a través de la explicabilidad; y (v) definir un consenso para la evaluación holística de las explicaciones generadas por las técnicas XAI permitiendo la comparación entre éstas. Además, para satisfacer dichos objetivos (respectivamente), presentamos una solución compuesta por: (i) un modelo de requisitos para la captura de los objetivos y las necesidades usuarios implicados en procesos de IA donde se incorpora XAI.; (ii) un modelo conceptual para la selección de técnicas XAI en la integración de la explicabilidad en los escenarios de IA; (iii) una derivación entre ambos modelos para llegar a la selección de técnicas de explicabilidad desde los requisitos XAI; (iv) un proceso para el análisis de la equidad en los datos gracias a la generación de explicabilidad en escenarios de IA basado en un algoritmo recursivo -incluyendo métricas de equidad para la evaluación de ésta-; y (v) presentamos una propuesta para la evaluación de la explicabilidad desde una perspectiva MDD alineada con los estándares ISO 25012 y 25024. Para validar las propuestas de esta tesis doctoral y verificar si se satisfacen o no los objetivos previamente definidos, se han aplicado en distintos casos de estudio en los que se destaca el proyecto de investigación BALLADEER enfocado en el diagnóstico y tratamiento del Trastorno del Déficit de Atención e Hiperactividad o TDAH a través de la IA. Como resultado, se ha podido comprobar como a través de los modelos presentados la aplicación de XAI ha llevado a diseñar interfaces explicables de manera exitosa. Por otra parte, el proceso presentado ha arrojado luz sobre cómo de sesgado son los modelos de ML aplicados en los distintos conjuntos de datos incluidos. Finalmente, la propuesta para evaluar XAI ha servido para poder comparar y concluir qué aproximación explicable funcionaba de manera más eficiente atendiendo a todas las dimensiones y los usuarios que se han implicado en el escenario de IA en cuestión. Así pues, se han satisfecho los objetivos establecidos a la vez que se han abierto líneas de investigación futura.
RUA. Repositorio Institucional de la Universidad de Alicante
Publicación Otros (Other)

A Shape Expression approach for assessing the quality of Linked Open Data in Libraries

  • Candela, Gustavo
  • Escobar Esteban, María Pilar
  • Sáez Fernández, María Dolores
  • Marco Such, Manuel
  • Candela, Gustavo|||/items/df3bee46-829c-44a2-acd0-49044a591ffa
  • Escobar Esteban, María Pilar|||/items/e08f4b8d-724a-444f-93d3-56f4e07864a7
  • Marco Such, Manuel|||/items/f781969d-dee8-453a-954b-d12cd9d165f8
  • Sáez Fernández, María Dolores|||/items/36aa1ebf-ac25-45ed-bced-6888fa74f638
Cultural heritage institutions are exploring Semantic Web technologies to publish and enrich their catalogues. Several initiatives, such as Labs, are based on the creative and innovative reuse of the materials published by cultural heritage institutions. In this way, quality has become a crucial aspect to identify and reuse a dataset for research. In this article, we propose a methodology to create Shape Expressions definitions in order to validate LOD datasets published by libraries. The methodology was then applied to four use cases based on datasets published by relevant institutions. It intends to encourage institutions to use ShEx to validate LOD datasets as well as to promote the reuse of LOD, made openly available by libraries.
RUA. Repositorio Institucional de la Universidad de Alicante
Publicación Otros (Other)

Evaluating CNN Methods for Epileptic Seizure Type Classification Using EEG Data

  • Rivera, Manuel J.
  • Sanchis, Javier
  • Corcho, Oscar
  • Teruel, Miguel A.
  • Trujillo, Juan
  • Sanchis, Javier|||/items/223d926e-7f4e-416b-89b9-2ee8648e6fcc
  • Teruel, Miguel A.|||/items/72e39554-945f-4f46-9aea-dbce9f65e43b
  • Trujillo, Juan|||/items/6b71a8e1-3f53-45b5-b548-767fac0d66aa
Epilepsy impacts around 6.38 per 1,000 people globally, presenting diagnostic challenges due to the complexity of seizures. Accurate classification of seizure types via Electroencephalogram (EEG) is critical for effective treatment and enhancing patient quality of life. However, the intricate characteristics of EEG data necessitate expert interpretation, a process that is both time-intensive and susceptible to human error. Recent advancements in Deep Learning (DL) have shown promise in EEG analysis, offering new avenues for seizure type classification. This study introduces two innovative deep learning architectures designed for seizure type classification within a seven-class framework: Network 1D Raw, which applies 1D Convolutions to Raw EEG signals, and Network 2D Conv, utilizing 2D convolutions on pre-computed spectrograms. Both architectures employ Separable Convolutions to enhance feature extraction efficiency, with the Network 1D Raw also incorporating a dilation rate technique for expanded analysis. Tested on the Temple University Hospital Seizure (TUSZ) dataset, these methods are evaluated using inter-patient 3-fold cross-validation. The Network 1D Raw achieved a weighted f1-score of 0.611 ± 0.037, while the Network 2D Conv reached 0.599±0.052, both surpassing existing benchmarks. The Network 1D Raw demonstrated superior classification for Absence Seizure (ABSZ), Focal Non-specific Seizures (FNSZ), and Generalized Non-specific Seizures types (GNSZ) with AUPRs of 0.9400±0.0500, 0.8100±0.0300, and 0.5500±0.1800, with Network 2D Conv excelling in FNSZ Seizures and GNSZ classes with AUPRs of 0.8100±0.1100 and 0.4000 ± 0.0200. Our DL models advance seizure type classification, blending efficiency with accuracy. Network 1D Raw’s compact design suits low-resource environments, aiding quick, precise diagnoses. Future work will focus on expanding dataset diversity, particularly for underrepresented seizure types, to further refine classification performance.
RUA. Repositorio Institucional de la Universidad de Alicante
Publicación Otros (Other)

Attention Robots: Juego serio para la ayuda de la detección del TDAH

  • Sanchis, Javier
  • Panagiotidis-Arrizabalaga, Alejandro
  • Maciá, Marina
  • Fernández de Córdoba, Alejandro
  • Teruel, Miguel A.
  • Trujillo, Juan
  • Sanchis, Javier|||/items/223d926e-7f4e-416b-89b9-2ee8648e6fcc
  • Panagiotidis-Arrizabalaga, Alejandro|||/items/638032a1-01c0-4559-a0b5-1b28e50d8786
  • Maciá, Marina|||/items/5e045f47-05d6-43ac-a061-4073cad75e70
  • Teruel, Miguel A.|||/items/72e39554-945f-4f46-9aea-dbce9f65e43b
  • Trujillo, Juan|||/items/6b71a8e1-3f53-45b5-b548-767fac0d66aa
El Trastorno por Déficit de Atención e Hiperactividad (TDAH) es un trastorno neuroevolutivo prevalente durante la infancia y la adolescencia caracterizado por síntomas generalizados de inatención, hiperactividad e impulsividad, afectando significativamente el funcionamiento diario. Los métodos diagnósticos tradicionales para el TDAH son extensos y requieren de muchos recursos. La incorporación de Juegos Serios basados en el diagnóstico y la intervención terapéutica del TDAH representa un avance revolucionario en los dominios de la salud mental y la neurociencia. Este estudio tiene como objetivo desarrollar un Juego Serio, llamado “Attention Robots”, para facilitar un sistema asistido por ordenador para el diagnóstico del TDAH. El juego recopila y almacena datos biométricos mediante el uso de un casco de electroencefalografía (EEG), una pulsera de actividad electrodermal (EDA) y un dispositivo de seguimiento ocular. Los sujetos participan en una dinámica que requiere identificar robot basados en la postura, mientras el sistema registra meticulosamente aciertos, comisiones y omisiones para analizar el enfoque y la atención. La efectividad de “Attention Robots” se evaluó a través de un análisis comparativo con el establecido Test de Atención D2. Se empleó un diseño experimental controlado, demostrando la validez del juego como herramienta de evaluación de la atención. Las métricas comparativas enfatizaron el tiempo de respuesta, la tasa de error y la consistencia atencional, alineándose estrechamente con los parámetros del Test D2. En conclusión, “Attention Robots” significa un avance notable en la gamificación de la evaluación del TDAH, destacando el potencial de las plataformas interactivas basadas en juegos para agilizar el proceso de diagnóstico del TDAH y apoyar las intervenciones terapéuticas. Mientras que el estudio actual se centra en los mecanismos centrales del juego en la medición de la atención, se anticipan mejoras futuras que involucran la integración de tecnologías de análisis biométrico con el objetivo refinar aún más la capacidad del juego para una evaluación precisa del TDAH.
RUA. Repositorio Institucional de la Universidad de Alicante
Publicación Otros (Other)

Estudio de perspectiva de género en carreras técnicas

  • Lavalle, Ana
  • Teruel, Miguel A.
  • Maté, Alejandro
  • Trujillo, Juan
  • Lavalle, Ana|||/items/bd9039d2-f3e3-4ad6-8797-8c6874286dcf
  • Maté, Alejandro|||/items/93f2089c-972b-4620-bfbe-c0e3cc4a6fa3
  • Teruel, Miguel A.|||/items/72e39554-945f-4f46-9aea-dbce9f65e43b
  • Trujillo, Juan|||/items/6b71a8e1-3f53-45b5-b548-767fac0d66aa
En estos tiempos donde la igualdad entre géneros es un asunto necesario de interés general, todavía quedan ámbitos en la educación superior en los que la presencia masculina es abrumadora. Hablamos de las carreras técnicas en general y de la Ingeniería Informática en particular, en la cual encontramos solamente un 16 % de presencia femenina. Este hecho nos ha hecho plantearnos el porqué de esta desigualdad, al cual hemos intentado dar respuesta mediante una encuesta realizada a una serie de estudiantes de carreras técnicas. Gracias a esta encuesta hemos sido capaces de entender de primera mano cuáles son las motivaciones y opiniones del estudiantado de carreras técnicas respecto a la perspectiva de género en las mismas. Además, hemos planteado una serie de propuestas de cara a fomentar la presencia femenina en esos estudios en base a los resultados de la encuesta.
RUA. Repositorio Institucional de la Universidad de Alicante
Publicación Otros (Other)

Architecture design of a reinforcement environment for learning sign languages

  • Naranjo-Zeledón, Luis
  • Chacón Rivas, Mario
  • Peral, Jesús
  • Ferrández, Antonio
  • Ferrández, Antonio|||/items/b77d3a3a-c05b-4aa1-921f-a0f269e50ba1
  • Peral, Jesús|||/items/c944ebfe-a23a-4bde-b779-e2054a62d1a7
Different fields such as linguistics, teaching, and computing have demonstrated special interest in the study of sign languages (SL). However, the processes of teaching and learning these languages turn complex since it is unusual to find people teaching these languages that are fluent in both SL and the native language of the students. The teachings from deaf individuals become unique. Nonetheless, it is important for the student to lean on supportive mechanisms while being in the process of learning an SL. Bidirectional communication between deaf and hearing people through SL is a hot topic to achieve a higher level of inclusion. However, all the processes that convey teaching and learning SL turn difficult and complex since it is unusual to find SL teachers that are fluent also in the native language of the students, making it harder to provide computer teaching tools for different SL. Moreover, the main aspects that a second language learner of an SL finds difficult are phonology, non-manual components, and the use of space (the latter two are specific to SL, not to spoken languages). This proposal appears to be the first of the kind to favor the Costa Rican Sign Language (LESCO, for its Spanish acronym), as well as any other SL. Our research focus stands on reinforcing the learning process of final-user hearing people through a modular architectural design of a learning environment, relying on the concept of phonological proximity within a graphical tool with a high degree of usability. The aim of incorporating phonological proximity is to assist individuals in learning signs with similar handshapes. This architecture separates the logic and processing aspects from those associated with the access and generation of data, which makes it portable to other SL in the future. The methodology used consisted of defining 26 phonological parameters (13 for each hand), thus characterizing each sign appropriately. Then, a similarity formula was applied to compare each pair of signs. With these pre-calculations, the tool displays each sign and its top ten most similar signs. A SUS usability test and an open qualitative question were applied, as well as a numerical evaluation to a group of learners, to validate the proposal. In order to reach our research aims, we have analyzed previous work on proposals for teaching tools meant for the student to practice SL, as well as previous work on the importance of phonological proximity in this teaching process. This previous work justifies the necessity of our proposal, whose benefits have been proved through the experimentation conducted by different users on the usability and usefulness of the tool. To meet these needs, homonymous words (signs with the same starting handshape) and paronyms (signs with highly similar handshape), have been included to explore their impact on learning. It allows the possibility to apply the same perspective of our existing line of research to other SL in the future.
RUA. Repositorio Institucional de la Universidad de Alicante
Publicación Otros (Other)

Developing an interpretable machine learning model for predicting COVID-19 patients deteriorating prior to intensive care unit admission using laboratory markers

  • Reina Reina, Alejandro
  • Barrera, Jose Manuel
  • Maté, Alejandro
  • Trujillo, Juan
  • Valdivieso, Bernardo
  • Gas, María-Eugenia
  • Maté, Alejandro|||/items/93f2089c-972b-4620-bfbe-c0e3cc4a6fa3
  • Reina Reina, Alejandro|||/items/3dec779b-9915-4612-ba90-565bb21ad2a0
  • Trujillo, Juan|||/items/6b71a8e1-3f53-45b5-b548-767fac0d66aa
Coronavirus disease (COVID-19) remains a significant global health challenge, prompting a transition from emergency response to comprehensive management strategies. Furthermore, the emergence of new variants of concern, such as BA.2.286, underscores the need for early detection and response to new variants, which continues to be a crucial strategy for mitigating the impact of COVID-19, especially among the vulnerable population. This study aims to anticipate patients requiring intensive care or facing elevated mortality risk throughout their COVID-19 infection while also identifying laboratory predictive markers for early diagnosis of patients. Therefore, haematological, biochemical, and demographic variables were retrospectively evaluated in 8,844 blood samples obtained from 2,935 patients before intensive care unit admission using an interpretable machine learning model. Feature selection techniques were applied using precision-recall measures to address data imbalance and evaluate the suitability of the different variables. The model was trained using stratified cross-validation with k=5 and internally validated, achieving an accuracy of 77.27%, sensitivity of 78.55%, and area under the receiver operating characteristic (AUC) of 0.85; successfully identifying patients at increased risk of severe progression. From a medical perspective, the most important features of the progression or severity of patients with COVID-19 were lactate dehydrogenase, age, red blood cell distribution standard deviation, neutrophils, and platelets, which align with findings from several prior investigations. In light of these insights, diagnostic processes can be significantly expedited through the use of laboratory tests, with a greater focus on key indicators. This strategic approach not only improves diagnostic efficiency but also extends its reach to a broader spectrum of patients. In addition, it allows healthcare professionals to take early preventive measures for those most at risk of adverse outcomes, thereby optimising patient care and prognosis.
RUA. Repositorio Institucional de la Universidad de Alicante
Publicación Otros (Other)

Data Analytics and Artificial Intelligence in the New Scenario of Data Spaces

  • Trujillo, Juan
  • Reina Reina, Alejandro
  • Candela, Gustavo
  • Candela, Gustavo|||/items/df3bee46-829c-44a2-acd0-49044a591ffa
  • Reina Reina, Alejandro|||/items/3dec779b-9915-4612-ba90-565bb21ad2a0
  • Trujillo, Juan|||/items/6b71a8e1-3f53-45b5-b548-767fac0d66aa
Data spaces present significant opportunities for organizations to collaborate and leverage decentralized, interoperable, and secure data for advanced analytics and Federated Machine Learning. However, challenges such as ensuring data quality, managing privacy, and integrating heterogeneous data formats and semantics remain critical. Addressing these challenges requires robust data governance, real-time quality assurance, and the adoption of AI-data integration tools to improving decision-making and generating value in dynamic business ecosystems.
RUA. Repositorio Institucional de la Universidad de Alicante
Publicación Otros (Other)

Conceptual modeling in the era of Big Data and Artificial Intelligence: Research topics and introduction to the special issue

  • Trujillo, Juan
  • Davis, Karen C.
  • Du, Xiaoyong
  • Damiani, Ernesto
  • Storey, Veda C.
  • Trujillo, Juan|||/items/6b71a8e1-3f53-45b5-b548-767fac0d66aa
Since the first version of the Entity–Relationship (ER) model proposed by Peter Chen over forty years ago, both the ER model and conceptual modeling activities have been key success factors for modeling computer-based systems. During the last decade, conceptual modeling has been recognized as an important research topic in academia, as well as a necessity for practitioners. However, there are many research challenges for conceptual modeling in contemporary applications such as Big Data, data-intensive applications, decision support systems, e-health applications, and ontologies. In addition, there remain challenges related to the traditional efforts associated with methodologies, tools, and theory development. Recently, novel research is uniting contributions from both the conceptual modeling area and the Artificial Intelligence discipline in two directions. The first one is efforts related to how conceptual modeling can aid in the design of Artificial Intelligence (AI) and Machine Learning (ML) algorithms. The second one is how Artificial Intelligence and Machine Learning can be applied in model-based solutions, such as model-based engineering, to infer and improve the generated models. For the first time in the history of Conceptual Modeling (ER) conferences, we encouraged the submission of papers based on AI and ML solutions in an attempt to highlight research from both communities. In this paper, we present some of important topics in current research in conceptual modeling. We introduce the selected best papers from the 37th International Conference on Conceptual Modeling (ER’18) held in Xi’an, China and summarize some of the valuable contributions made based on the discussions of these papers. We conclude with suggestions for continued research.
RUA. Repositorio Institucional de la Universidad de Alicante
Publicación Otros (Other)

Machine learning model from a Spanish cohort for prediction of SARS-COV-2 mortality risk and critical patients

  • Reina Reina, Alejandro
  • Barrera, Jose Manuel
  • Valdivieso, Bernardo
  • Gas, María-Eugenia
  • Maté, Alejandro
  • Trujillo, Juan
  • Maté, Alejandro|||/items/93f2089c-972b-4620-bfbe-c0e3cc4a6fa3
  • Reina Reina, Alejandro|||/items/3dec779b-9915-4612-ba90-565bb21ad2a0
  • Trujillo, Juan|||/items/6b71a8e1-3f53-45b5-b548-767fac0d66aa
Patients affected by SARS-COV-2 have collapsed healthcare systems around the world. Consequently, different challenges arise regarding the prediction of hospital needs, optimization of resources, diagnostic triage tools and patient evolution, as well as tools that allow us to analyze which are the factors that determine the severity of patients. Currently, it is widely accepted that one of the problems since the pandemic appeared was to detect (i) who patients were about to need Intensive Care Unit (ICU) and (ii) who ones were about not overcome the disease. These critical patients collapsed Hospitals to the point that many surgeries around the world had to be cancelled. Therefore, the aim of this paper is to provide a Machine Learning (ML) model that helps us to prevent when a patient is about to be critical. Although we are in the era of data, regarding the SARS-COV-2 patients, there are currently few tools and solutions that help medical professionals to predict the evolution of patients in order to improve their treatment and the needs of critical resources at hospitals. Moreover, most of these tools have been created from small populations and/or Chinese populations, which carries a high risk of bias. In this paper, we present a model, based on ML techniques, based on 5378 Spanish patients’ data from which a quality cohort of 1201 was extracted to train the model. Our model is capable of predicting the probability of death of patients with SARS-COV-2 based on age, sex and comorbidities of the patient. It also allows what-if analysis, with the inclusion of comorbidities that the patient may develop during the SARS-COV-2 infection. For the training of the model, we have followed an agnostic approach. We explored all the active comorbidities during the SARS-COV-2 infection of the patients with the objective that the model weights the effect of each comorbidity on the patient’s evolution according to the data available. The model has been validated by using stratified cross-validation with k = 5 to prevent class imbalance. We obtained robust results, presenting a high hit rate, with 84.16% accuracy, 83.33% sensitivity, and an Area Under the Curve (AUC) of 0.871. The main advantage of our model, in addition to its high success rate, is that it can be used with medical records in order to predict their diagnosis, allowing the critical population to be identified in advance. Furthermore, it uses the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD 9-CM) standard. In this sense, we should also emphasize that those hospitals using other encodings can add an intermediate layer business to business (B2B) with the aim of making transformations to the same international format.
RUA. Repositorio Institucional de la Universidad de Alicante
Publicación Otros (Other)

Security policies by design in NoSQL document databases

  • Blanco Bueno, Carlos
  • García-Saiz, Diego
  • García Rosado, David
  • Santos Olmo, Antonio
  • Peral, Jesús
  • Maté, Alejandro
  • Trujillo, Juan
  • Fernández-Medina Patón, Eduardo
  • Maté, Alejandro|||/items/93f2089c-972b-4620-bfbe-c0e3cc4a6fa3
  • Peral, Jesús|||/items/c944ebfe-a23a-4bde-b779-e2054a62d1a7
  • Trujillo, Juan|||/items/6b71a8e1-3f53-45b5-b548-767fac0d66aa
The importance of data security is currently increasing owing to the number of data transactions that are continuously taking place. Large amounts of data are generated, stored, modified and transferred every second, signifying that databases require an appropriate capacity, control and protection that will enable them to maintain a secure environment for so much data. Big Data is becoming a prominent trend in our society, and increasing amounts of data, including sensitive and personal information, are being loaded into NoSQL and other Big Data technologies for analysis and processing. However, current security approaches do not take into account the special characteristics of these technologies, leaving sensitive and personal data unprotected and consequently risking considerable financial losses and brand damage. In this paper, we focus on NoSQL document databases and present a proposal for the design and implementation of security policies in this type of databases. We first follow the concept of security by design in order to propose a metamodel that allows the specification of both the structure and the security policies required for document databases. We also define an implementation model by analysing the implementation features provided by a specific NoSQL document database management system (MongoDB). Having obtained the design and implementation models, we follow the model-driven development philosophy and propose a set of transformation rules that allow the automatic generation of the final implementation of security policies. We additionally provide a technological solution in which the Eclipse Modelling Framework environment is employed in order to implement both the design metamodel (Emfatic) and the transformations (Epsilon, EGL). Finally, we apply the proposed framework to a case study carried out in the airport domain. This proposal, in addition to saving development time and costs, generates more robust solutions by considering security by design. This, therefore, abstracting the designer from both specific aspects of the target tool and having to choose the best strategies for the implementation of security policies.
RUA. Repositorio Institucional de la Universidad de Alicante
Publicación Otros (Other)

Computer-Aided Diagnostic Tools based on Encephalography and Deep Learning for the detection of mental issues: case of Attention Deficit Hyperactivity Disorder

  • Sanchis, Javier
  • Sanchis, Javier|||/items/223d926e-7f4e-416b-89b9-2ee8648e6fcc
Attention-Deficit/Hyperactivity Disorder (ADHD) is a prevalent mental health issue that significantly affects daily functioning and quality of life, especially in children and adolescents. Traditional methods of diagnosing ADHD are based on the DSM-5 criteria and tend to be subjective, relying heavily on symptoms reported by patients or their parents. In addition, these methods are supported by tests that can give inaccurate answers, affecting the accuracy of the diagnosis. The process can also be lengthy, as the DSM-5 requires at least 6 months of symptoms for diagnosis confirmation. These challenges highlight the need to implement more objective and less time-consuming diagnostic tools. The development of neural networks for ADHD diagnosis using biometric signals, such as electroencephalography (EEG), can help clinicians address these challenges. By employing a signal that directly reflects brain activity, it is possible to provide more objective diagnoses based on measurable data. Once the EEG signal is obtained, the computer system can generate results in a matter of seconds, as the neural networks have been pre-trained. Moreover, these networks can be continuously updated with new data from diagnosed subjects, thus improving the accuracy of the diagnosis. Finally, such systems could also be widely implemented, for example, in schools, allowing a larger number of potential ADHD cases to be detected for further evaluation by a specialist. It should be noted that these systems are not intended to replace physicians, but rather to help them detect more cases, thus enabling them to treat a larger number of people. Although several works have been developed in the field of neural networks and EEG, the results can be improved and the tools can be made more reliable and replicable. Another technology that could help address the inherent limitations of traditional diagnostic methods is serious games. These gamified systems can capture real-time data on the responses and decision-making abilities of players, providing a more objective assessment of symptoms such as inattention and impulsivity. In addition, video games may not be perceived as a form of treatment or an imposition by parents or teachers, potentially making them more appealing to children. Furthermore, according to the DSM-5, no specific cause for ADHD has been identified, which has generated interest in studying the underlying neural mechanisms. However, studies on the brain regions associated with ADHD show heterogeneous results, pointing to different hemispheres and lobes. This divergence underlines the complexity of ADHD and highlights the need for further research. Finally, with a better understanding of the origin of ADHD, more cost-effective and specialized EEG signal capture devices could be developed. To address the aforementioned challenges, the main goal of this doctoral thesis is to explore the development and implementation of computer-aided systems, leveraging computer engineering and artificial intelligence, to support the diagnosis and enhance the understanding of ADHD. From this objective, three lines of research are defined: (i) developing neural networks using EEG signals to assist in the diagnosis of ADHD, (ii) identifying the brain regions that best characterize ADHD, and (iii) creating serious games as diagnostic tools and biometric data collection systems to detect ADHD in an objective and reliable manner. First, a neural network was developed that achieved an accuracy of 88.42%, outperforming existing models. The hyperparameter tuning of this network also reduced the training time from 15.5 days to 1.3 days, without compromising performance. This tuning enabled the development of new architectures by focusing on key elements of the neural network, leading to a second, more accurate network for diagnosis, achieving 94.87% accuracy. Secondly, this work identified key brain regions relevant to ADHD diagnosis by analyzing various subsets of EEG channels. The results showed that the Frontal Lobe and Left Hemisphere provided significant information for classifying ADHD, achieving f1-scores of 80.81% and 80.56%, respectively. Feature selection methods, such as Backward Stepwise Selection, improved classification performance to an f1-score of 82.81%. Finally, the Serious Game “Attention Slackline” was designed to contribute to the assessment of attention with objective and systematic measures. During the game, intrinsic metrics such as hits, omissions, and commissions were collected. In addition, biometric data such as EEG, electrodermal activity, and eye tracking were recorded, providing a high-quality dataset for future research. To test the feasibility of the game as a tool for measuring attention, a protocol was defined in collaboration with a team of psychologists. The aim was to test whether the results obtained from the metrics obtained during the game correlate with those obtained from the well-known D2 attention test. As we obtained positive correlations, it is confirmed that “Attention Slackline” is a valid tool for measuring attention. The findings of this thesis indicate that integrating AI techniques and serious games into ADHD diagnostics can lead to more accurate, efficient, and engaging assessments. Future work will focus on exploring other DL architectures, incorporating multimodal data, developing personalized models, and ensuring the explainability of these AI systems to facilitate their integration into clinical practice.
RUA. Repositorio Institucional de la Universidad de Alicante
Publicación Otros (Other)

Machine Learning: Requisitos y nuevas técnicas para la aplicación en entornos industriales e Internet of Things

  • Barrera, Jose Manuel
Con la conversión de la industria tradicional a industria 4.0, la inteligencia artificial (IA), Internet de las Cosas (IoT), el Machine Learning (ML) y la calidad de los datos (DQ) adquieren nuevas dimensiones y una notoria relevancia en el campo de la industria. Uno de los usos más solicitados de la IA en el campo de la industria es la búsqueda de la mejora de la rentabilidad, ya sea aumentando la producción, o disminuyendo los costes del propio proceso industrial. En esta tesis se abordan los dos objetivos: para la optimización de producción, se presenta un sistema para la cuantificación de energía generada para instalaciones solares fotovoltaicas. Este modelo está basado en Open Data provenientes de satélites, sensores IoT y Redes Neuronales Artificiales (ANN). Así, facilitamos a los lectores la información necesaria para decidir cuánto deberían invertir en una ubicación en concreto, en función de la producción energética deseada. En comparación con las propuestas más avanzadas, nuestra solución proporciona una capa de abstracción centrada en la producción de energía, en lugar de en los datos de radiación, y puede entrenarse y adaptarse a diferentes ubicaciones utilizando Open Data. Por otra parte, para la disminución de costes del propio proceso industrial, se presenta un modelo de ML basado en autoencoders que permite la Detección y el Diagnosis de Errores (FDD) y acorta las interrupciones del proceso productivo. El enfoque presentado explota los datos generados por el proceso industrial y entrena una arquitectura basada en ML, combinando varios algoritmos con autoencoders y ventanas deslizantes. La solución presentada ayuda a detectar precozmente las averías y se ha sido probada con datos reales procedentes de una instalación con una turbina de cogeneración de energía eléctrica. Además, se destaca que, aunque el ejemplo usado para nuestro enfoque utiliza una instalación industrial de una turbina de gas, éste puede adaptarse a otros problemas de FDD en otros procesos industriales que podrían beneficiarse de las ventajas mencionadas. Sin embargo, durante la realización del proyecto se han encontrado dos dificultades adicionales: que no existe una metodología establecida para la captación de requisitos para proyectos de ML; y que no existe información adecuada y suficiente sobre el efecto de las desviaciones de la Calidad de lo Datos (DQ) en dichos modelos ML. Por consiguiente, en esta tesis se presentan dos soluciones adicionales para dichos problemas. Para solventar la captación de requisitos en proyectos de ML, se presenta una particularización del modelo de requisitos iStar (i*). Mediante esta particularización, los recién llegados y los no expertos en el campo del ML pueden seguir una metodología que les guiará por el camino correcto a seguir, evitando modelos no válidos. El marco i* es un popular lenguaje de modelado para capturar el entorno y los requisitos de un sistema. Sin embargo, se ha construido sobre una capa de abstracción muy alta, y sus usos en un campo específico dependen en gran medida de la experiencia del diseñador. La propuesta presentada resuelve esto: especifica el marco i*, cubriendo las principales lagunas entre el ML y el modelado conceptual. Así, proporcionamos una línea de base adecuada a seguir que captura los requisitos y cumple las numerosas restricciones del campo del ML. Además, se presenta una guía basada en preguntas para aplicar dicha propuesta, y se aplica la metodología descrita en el proyecto de la turbina de gas descrito previamente. Así, se puede ver su viabilidad y cómo su uso filtra diseños no válidos. Respecto al punto de ver la cuantificación del efecto de DQ en los proyectos de ML, en esta tesis se presenta un enfoque sistemático, basado en la norma ISO 25012, para estimar el impacto de la degradación de la DQ en diferentes algoritmos, con el fin de cuantificar dicho efecto en la salida de un modelo de ML de una instalación real (la de la turbina de gas de cogeneración). Se debe destacar que en una instalación real puede haber fuertes restricciones de tiempo y espacio, y la limpieza de datos como tal no siempre es posible, por lo que el modelo de ML debe lidiar de con dichos problemas de DQ. Para ello, se ha definido una metodología mediante la cual, se contamina progresivamente los datos para disminuir dos características de D, accuracy y currentness. Como resumen, la tesis se centra en 4 puntos: optimización de un proceso industrial de generación solar fotovoltaica, mejora en el mantenimiento industrial mediante mantenimiento predictivo en una instalación real, metodología para la captación de requisitos en un proyecto de ML, y efectos de DQ sobre las salidas de los modelos de ML en un entorno industrial. Así pues, esta tesis mejora de manera integral distintos aspectos relacionados con el ML, el IoT y las instalaciones industriales.
RUA. Repositorio Institucional de la Universidad de Alicante
Publicación Otros (Other)

An automatic data quality approach to assess semantic data from cultural heritage institutions

  • Candela, Gustavo
  • Candela, Gustavo|||/items/df3bee46-829c-44a2-acd0-49044a591ffa
In recent years, cultural heritage institutions have been exploring the benefits of applying Linked Open Data to their catalogs and digital materials. Innovative and creative methods have emerged to publish and reuse digital contents to promote computational access, such as the concepts of Labs and Collections as Data. Data quality has become a requirement for researchers and training methods based on artificial intelligence and machine learning. This article explores how the quality of Linked Open Data made available by cultural heritage institutions can be automatically assessed. The results obtained can be useful for other institutions who wish to publish and assess their collections.
RUA. Repositorio Institucional de la Universidad de Alicante
Publicación Otros (Other)

Refining ADHD diagnosis with EEG: The impact of preprocessing and temporal segmentation on classification accuracy

  • García-Ponsoda, Sandra
  • Maté, Alejandro
  • Trujillo, Juan
  • García-Ponsoda, Sandra|||/items/b1adddd3-e77f-4ee6-bc2e-7b00acbae17c
  • Maté, Alejandro|||/items/93f2089c-972b-4620-bfbe-c0e3cc4a6fa3
  • Trujillo, Juan|||/items/6b71a8e1-3f53-45b5-b548-767fac0d66aa
Background: EEG signals are commonly used in ADHD diagnosis, but they are often affected by noise and artifacts. Effective preprocessing and segmentation methods can significantly enhance the accuracy and reliability of ADHD classification. Methods: We applied filtering, ASR, and ICA preprocessing techniques to EEG data from children with ADHD and neurotypical controls. The EEG recordings were segmented, and features were extracted and selected based on statistical significance. Classification was performed using various EEG segments and channels with Machine Learning models (SVM, KNN, and XGBoost) to identify the most effective combinations for accurate ADHD diagnosis. Results: Our findings show that models trained on later EEG segments achieved significantly higher accuracy, indicating the potential role of cognitive fatigue in distinguishing ADHD. The highest classification accuracy (86.1%) was achieved using data from the P3, P4, and C3 channels, with key features such as Kurtosis, Katz fractal dimension, and power spectrums in the Delta, Theta, and Alpha bands contributing to the results. Conclusion: This study highlights the importance of preprocessing and segmentation in improving the reliability of ADHD diagnosis through EEG. The results suggest that further research on cognitive fatigue and segmentation could enhance diagnostic accuracy in ADHD patients.
RUA. Repositorio Institucional de la Universidad de Alicante
Publicación Otros (Other)

Enabling the Application of Graph Neural Networks on Graphs With Unknown Connectivity

  • García Carrasco, Jorge
  • Maté, Alejandro
  • Trujillo, Juan
  • Maté, Alejandro|||/items/93f2089c-972b-4620-bfbe-c0e3cc4a6fa3
  • Trujillo, Juan|||/items/6b71a8e1-3f53-45b5-b548-767fac0d66aa
  • García Carrasco, Jorge|||/items/cf588298-eb32-4b82-9b26-794828afb69c
Graph Neural Networks (GNNs) have proven to be reliable methods for working with graph-structured data. However, it is common to find graphs with partially or fully inaccessible connectivity patterns, hindering the direct application of GNNs to the task at hand. To tackle this problem, several Graph Structure Learning (GSL) methods have been proposed, with the objective of jointly optimizing both the graph structure and the GNN model by adding loss terms that enforce desired graph properties. These properties, such as sparseness and connectivity of similar nodes, can have a drastic impact on the performance of a GNN. However, current methods offer little control on the desired degree of sparseness, which may lead to non-optimal connectivity and reduced efficiency. In this paper, we propose a new method called Adaptative Sparsification Graph Learning (ASGL), which enables fine-grained, linear control over the total number of edges in the resulting learned graph via a novel perturbation-based loss term. ASGL not only provides flexibility in sparsity control but also improves both accuracy and computational efficiency, outperforming state-of-the-art methods in most benchmarks. We demonstrate its robustness through extensive experiments and highlight how adjusting sparsity enables optimizing the trade-off between accuracy, complexity, and interpretability.
RUA. Repositorio Institucional de la Universidad de Alicante
Publicación Otros (Other)

A methodology for the systematic design of storytelling dashboards applied to Industry 4.0

  • Lavalle, Ana
  • Maté, Alejandro
  • Santos, Maribel Yasmina
  • Guimarães, Pedro
  • Trujillo, Juan
  • Santos, Antonina
  • Lavalle, Ana|||/items/bd9039d2-f3e3-4ad6-8797-8c6874286dcf
  • Maté, Alejandro|||/items/93f2089c-972b-4620-bfbe-c0e3cc4a6fa3
  • Trujillo, Juan|||/items/6b71a8e1-3f53-45b5-b548-767fac0d66aa
Dashboards are popular tools for presenting key insights to decision-makers by translating large volumes of data into clear information. However, while individual visualizations may effectively answer specific questions, they often fail to connect in a way that conveys the overall narrative, leaving decision-makers without a cohesive understanding of the area under analysis. This paper presents a novel methodology for the systematic design of holistic dashboards, moving from analytical requirements to storytelling dashboards. Our approach ensures that all visualizations are aligned with the analytical goals of decision-makers. It includes several key steps: capturing analytical requirements through the i* framework; structuring and refining these requirements into a tree model to reflect the decision-maker’s mental analysis; identifying and preparing relevant data; capturing the key concepts and relationships for the composition of the cohesive storytelling dashboard through a novel storytelling conceptual model; finally, implementing and integrating the visualizations into the dashboard, ensuring coherence and alignment with the decision-maker’s needs. Our methodology has been applied in real-world industrial environments. We evaluated its impact through a controlled experiment. The findings show that storytelling dashboards significantly improve data interpretation, reduce misinterpretations, and enhance the overall user experience compared to traditional dashboards.
RUA. Repositorio Institucional de la Universidad de Alicante
Publicación Otros (Other)

CuentosIE: can a chatbot about “tales with a message” help to teach emotional intelligence?

  • Ferrández, Antonio
  • Lavigne Cerván, Rocio
  • Peral, Jesús
  • Navarro Soria, Ignasi
  • Lloret, Ángel
  • Gil, David
  • Rocamora-Rodríguez, María C.
  • Ferrández, Antonio|||/items/b77d3a3a-c05b-4aa1-921f-a0f269e50ba1
  • Gil, David|||/items/90b814a5-7a43-4cc4-bd20-9a77d5ff0f08
  • Lloret, Ángel|||/items/f87b9fb9-6755-4011-8dc5-af571153e1f5
  • Navarro Soria, Ignasi|||/items/1c7c4096-530f-4487-8b9c-e472938c5df6
  • Peral, Jesús|||/items/c944ebfe-a23a-4bde-b779-e2054a62d1a7
  • Rocamora-Rodríguez, María C.|||/items/860042c9-93c8-45e7-b00f-98172f389073
In this article, we present CuentosIE (TalesEI: chatbot of tales with a message to develop Emotional Intelligence), an educational chatbot on emotions that also provides teachers and psychologists with a tool to monitor their students/patients through indicators and data compiled by CuentosIE. The use of “tales with a message” is justified by their simplicity and easy understanding, thanks to their moral or associated metaphors. The main contributions of CuentosIE are the selection, collection, and classification of a set of highly specialized tales, as well as the provision of tools (searching, reading comprehension, chatting, recommending, and classifying) that are useful for both educating users about emotions and monitoring their emotional development. The preliminary evaluation of the tool has obtained encouraging results, which provides an affirmative answer to the question posed in the title of the article.
RUA. Repositorio Institucional de la Universidad de Alicante
Publicación Otros (Other)

An extension of iStar for Machine Learning requirements by following the PRISE methodology

  • Barrera, Jose Manuel
  • Reina Reina, Alejandro
  • Lavalle, Ana
  • Maté, Alejandro
  • Trujillo, Juan
  • Lavalle, Ana|||/items/bd9039d2-f3e3-4ad6-8797-8c6874286dcf
  • Maté, Alejandro|||/items/93f2089c-972b-4620-bfbe-c0e3cc4a6fa3
  • Reina Reina, Alejandro|||/items/3dec779b-9915-4612-ba90-565bb21ad2a0
  • Trujillo, Juan|||/items/6b71a8e1-3f53-45b5-b548-767fac0d66aa
The rise of Artificial Intelligence (AI) and Deep Learning has led to Machine Learning (ML) becoming a common practice in academia and enterprise. However, a successful ML project requires deep domain knowledge as well as expertise in a plethora of algorithms and data processing techniques. This leads to a stronger dependency and need for communication between developers and stakeholders where numerous requirements come into play. More specifically, in addition to functional requirements such as the output of the model (e.g. classification, clustering or regression), ML projects need to pay special attention to a number of non-functional and quality aspects particular to ML. These include explainability, noise robustness or equity among others. Failure to identify and consider these aspects will lead to inadequate algorithm selection and the failure of the project. In this sense, capturing ML requirements becomes critical. Unfortunately, there is currently an absence of ML requirements modeling approaches. Therefore, in this paper we present the first i* extension for capturing ML requirements and apply it to two real-world projects. Our study covers two main objectives for ML requirements: (i) allows domain experts to specify objectives and quality aspects to be met by the ML solution, and (ii) facilitates the selection and justification of the most adequate ML approaches. Our case studies show that our work enables better ML algorithm selection, preprocessing implementation tailored to each algorithm, and aids in identifying missing data. In addition, they also demonstrate the flexibility of our study to adapt to different domains.
RUA. Repositorio Institucional de la Universidad de Alicante
Publicación Otros (Other)

Improving security in NoSQL document databases through model-driven modernization

  • Maté, Alejandro
  • Peral, Jesús
  • Trujillo, Juan
  • Blanco Bueno, Carlos
  • García-Saiz, Diego
  • Fernández-Medina Patón, Eduardo
  • Maté, Alejandro|||/items/93f2089c-972b-4620-bfbe-c0e3cc4a6fa3
  • Peral, Jesús|||/items/c944ebfe-a23a-4bde-b779-e2054a62d1a7
  • Trujillo, Juan|||/items/6b71a8e1-3f53-45b5-b548-767fac0d66aa
NoSQL technologies have become a common component in many information systems and software applications. These technologies are focused on performance, enabling scalable processing of large volumes of structured and unstructured data. Unfortunately, most developments over NoSQL technologies consider security as an afterthought, putting at risk personal data of individuals and potentially causing severe economic loses as well as reputation crisis. In order to avoid these situations, companies require an approach that introduces security mechanisms into their systems without scrapping already in-place solutions to restart all over again the design process. Therefore, in this paper we propose the first modernization approach for introducing security in NoSQL databases, focusing on access control and thereby improving the security of their associated information systems and applications. Our approach analyzes the existing NoSQL solution of the organization, using a domain ontology to detect sensitive information and creating a conceptual model of the database. Together with this model, a series of security issues related to access control are listed, allowing database designers to identify the security mechanisms that must be incorporated into their existing solution. For each security issue, our approach automatically generates a proposed solution, consisting of a combination of privilege modifications, new roles and views to improve access control. In order to test our approach, we apply our process to a medical database implemented using the popular document-oriented NoSQL database, MongoDB. The great advantages of our approach are that: (1) it takes into account the context of the system thanks to the introduction of domain ontologies, (2) it helps to avoid missing critical access control issues since the analysis is performed automatically, (3) it reduces the effort and costs of the modernization process thanks to the automated steps in the process, (4) it can be used with different NoSQL document-based technologies in a successful way by adjusting the metamodel, and (5) it is lined up with known standards, hence allowing the application of guidelines and best practices.
RUA. Repositorio Institucional de la Universidad de Alicante
Publicación Otros (Other)

A modeling approach for designing explainable Artificial Intelligence

  • Navarro, Álvaro
  • Lavalle, Ana
  • Maté, Alejandro
  • Trujillo, Juan
  • Lavalle, Ana|||/items/bd9039d2-f3e3-4ad6-8797-8c6874286dcf
  • Maté, Alejandro|||/items/93f2089c-972b-4620-bfbe-c0e3cc4a6fa3
  • Trujillo, Juan|||/items/6b71a8e1-3f53-45b5-b548-767fac0d66aa
  • Navarro, Álvaro|||/items/7ee415a4-651b-46a5-aeb9-32f9c9931cf8
EXplainable Artificial Intelligence (XAI) has become one of the most important and complex issues to address in the current Artificial Intelligence (AI) field. This topic is the key point to foster the adoption of the AI solutions, providing reliability in the AI systems. There are multiple XAI techniques, with its own characteristics such as the technical level or confidence required to make use of them. In this context, modeling approaches can aid in managing the complexity and describe the implementation alternatives available. However, there is a lack of these modeling approaches that help the XAI developers to select and apply the most adequate XAI techniques in the specific scenarios. Motivated by this issue, we aim to present a modeling approach to support XAI developers in the XAI processes. Then, we propose a conceptual model, which is represented in a Unified Modeling Language (UML) class diagram, to capture the key XAI elements. The main advantage of our proposed conceptual model is that it takes into consideration: (i) how the explanations should be generated; (ii) how these explanations will be combined and supported; and (iii) what explainable interfaces, where these explanations are included, will be generated and presented to the end-users. Finally, to test the applicability of our proposal, we have exemplified it through a real project focused on diagnosing the Attention-Deficit/Hyperactivity Disorder (ADHD).
RUA. Repositorio Institucional de la Universidad de Alicante
Publicación Otros (Other)

An ontology-based secure design framework for graph-based databases

  • Paneque, Manuel
  • Roldán-García, María del Mar
  • Blanco Bueno, Carlos
  • Maté, Alejandro
  • Rosado, David G.
  • Trujillo, Juan
  • Maté, Alejandro|||/items/93f2089c-972b-4620-bfbe-c0e3cc4a6fa3
  • Trujillo, Juan|||/items/6b71a8e1-3f53-45b5-b548-767fac0d66aa
Graph-based databases are concerned with performance and flexibility. Most of the existing approaches used to design secure NoSQL databases are limited to the final implementation stage, and do not involve the design of security and access control issues at higher abstraction levels. Ensuring security and access control for Graph-based databases is difficult, as each approach differs significantly depending on the technology employed. In this paper, we propose the first technology-ascetic framework with which to design secure Graph-based databases. Our proposal raises the abstraction level by using ontologies to simultaneously model database and security requirements together. This is supported by the TITAN framework, which facilitates the way in which both aspects are dealt with. The great advantages of our approach are, therefore, that it: allows database designers to focus on the simultaneous protection of security and data while ignoring the implementation details; facilitates the secure design and rapid migration of security rules by deriving specific security measures for each underlying technology, and enables database designers to employ ontology reasoning in order to verify whether the security rules are consistent. We show the applicability of our proposal by applying it to a case study based on a hospital data access control.
RUA. Repositorio Institucional de la Universidad de Alicante
Publicación Otros (Other)

Análisis y predicción de delincuencia en ciudades

  • Torre Vicedo, Santiago de la
La delincuencia es un problema que persiste en las ciudades modernas, y en concreto este TFG se centrará en los sucesos de tiroteos en la ciudad de Nueva York, con la finalidad de comprender sus patrones y tendencias. El objetivo de este TFG es llevar a cabo la integración de datos provenientes de diversas fuentes, con el propósito de procesar y transformar dichos datos en información que pueda ser posteriormente analizada. La integración de estos datos permitirá obtener una visión más amplia de la delincuencia en las ciudades, y por tanto, contribuirá con una comprensión más completa de estos sucesos. Además, se busca complementar el análisis con una serie de predicciones. Mediante el uso de técnicas de análisis predictivo, se buscará identificar patrones y correlaciones en los datos históricos. Estas predicciones permitirán obtener una visión de futuro y contribuir a la toma de decisiones para estrategias de prevención de delitos. Este campo de estudio posee cierta relevancia debido a su impacto en la seguridad ciudadana y seguridad de las ciudades. Al comprender ciertos factores que intervienen en los sucesos de tiroteos, se pueden diseñar medidas más efectivas y focalizadas en áreas específicas.
RUA. Repositorio Institucional de la Universidad de Alicante
Publicación Otros (Other)

Fault detection and diagnosis for industrial processes based on clustering and autoencoders: a case of gas turbines

  • Barrera, Jose Manuel
  • Reina Reina, Alejandro
  • Maté, Alejandro
  • Trujillo, Juan
  • Maté, Alejandro|||/items/93f2089c-972b-4620-bfbe-c0e3cc4a6fa3
  • Reina Reina, Alejandro|||/items/3dec779b-9915-4612-ba90-565bb21ad2a0
  • Trujillo, Juan|||/items/6b71a8e1-3f53-45b5-b548-767fac0d66aa
Industrial machinery maintenance constitutes an important part of the manufacturing company’s budget. Fault Detection and Diagnosis (henceforth referenced as FDD) plays a key role on maintenance, since it allows for shorter maintenance times and, in the long run, to train predictive maintenance algorithms. The impact of proper maintenance is reflected on an especially costly type of industrial machine: gas turbines. These devices are complex, large pieces of machinery that cause considerable service disruption when downtime occurs. In an effort to shorten these service disruptions and establish the basis for the development of predictive maintenance, we present in this paper an approach to FDD of industrial machinery, such as gas turbines. Our approach exploits the data generated by industrial machinery to train a machine-learning based architecture, combining several algorithms with autoencoders and sliding windows. Our proposed solution helps to achieve early malfunctioning detection and has been tested using real data from real working environments. In order to build our solution, first, we analyze the behavior of the gas turbine from a mathematical point of view. Then, we develop an architecture that is capable of detecting when the gas turbine presents an abnormal behavior. The great advantage of our proposal is that (i) does not require existing disruption data, which can be difficult to obtain, (ii) is not limited to processes with specific time windows, and (iii) provides crucial information in real time to the monitoring staff, generating valuable data for further predictive maintenance. It is worth highlighting that although we exemplify our approach using gas turbines, our approach can be tailored to other FDD problems in complex industrial processes with variable duration that could benefit from the aforementioned advantages.
RUA. Repositorio Institucional de la Universidad de Alicante
Publicación Otros (Other)

A Data Analytics Methodology to Visually Analyze the impact of Bias and Rebalancing

  • Lavalle, Ana
  • Maté, Alejandro
  • Trujillo, Juan
  • Teruel, Miguel A.
  • Lavalle, Ana|||/items/bd9039d2-f3e3-4ad6-8797-8c6874286dcf
  • Maté, Alejandro|||/items/93f2089c-972b-4620-bfbe-c0e3cc4a6fa3
  • Teruel, Miguel A.|||/items/72e39554-945f-4f46-9aea-dbce9f65e43b
  • Trujillo, Juan|||/items/6b71a8e1-3f53-45b5-b548-767fac0d66aa
Data Analytics have become a key component of many business processes which influence several aspects of our daily life. Indeed, any misinterpretation or flaw in the outputs of Data Analytics results can cause significant damage, specialy when dealing with one of the most often overlooked issues, namely the unaware use of biased data. When data bias goes unadverted, it warps the meaning of data, having a devastating effect on Data Analytics results. Although it is widely argued that the most common manner to deal with data bias is to rebalance biased datasets, it is not an aseptic transformation, leading to several potentially undesired side-effects that will probably harm the result of data analyses. Therefore, in order to analyze the underlying bias in datasets, in this work we present (i) a comprehensive methodology based on visualization techniques, which assists users in the definition of their analytical requirements to detect and visually represent the data bias automatically helping them to find out whether it is appropriate to artificially rebalance their dataset or not; (ii) a novel metamodel for visually representing bias; (iii) a motivating real-world running example used to analyze the impact of bias in Data Analytics and (iv) an assessment of the improvements introduced by our proposal through a complete real-world case study by using a Fire Department Calls for Service dataset, thus demonstrating that rebalancing datasets is not always the best option. It is crucial to study the context where the decisions are going to be taken. Moreover, it is also important to do a pre-analysis with the aim of knowing the nature of the datasets and how biased they are.