Persona: Pastor Vargas, Rafael
Cargando...
Dirección de correo electrónico
rpastor@scc.uned.es
ORCID
0000-0002-4089-9538
Fecha de nacimiento
Proyectos de investigación
Unidades organizativas
Puesto de trabajo
Apellidos
Pastor Vargas
Nombre de pila
Rafael
Nombre
37 resultados
Resultados de la búsqueda
Mostrando 1 - 10 de 37
Publicación EVI-LINHD, a virtual research environment for the Spanish speaking community(Oxford University Press, 2017-12) González-Blanco García, Elena; Rio Riande, Gimena del; Díez Platas, María Luisa; Olmo, Álvaro del; Urízar, Miguel; Martínez Cantón, Clara Isabel; Ros Muñoz, Salvador; Pastor Vargas, Rafael; Robles Gómez, Antonio; Caminero Herráez, Agustín CarlosLaboratorio de Innovación en Humanidades Digitales (UNED) has developed Entorno Virtual de Investigación del Laboratorio de Innovación en Humanidades Digitales (EVI-LINHD), the first virtual research environment devoted mainly to Spanish speakers interested in digital scholarly edition. EVI-LINHD combines different open-source software for developing a complete digital project: (1) a Webbased application markup tool—TEIscribe—combined with an eXistdb solution and a TEIPublisher platform, (2) Omeka for digital libraries, and (3) WordPress for simple Web pages. All these instances are linked to a local installation of the LINDAT/Common Language Resources and Technology Infrastructure (CLARIN) digital repository. LINDAT/CLARIN allows EVI-LINHD users to have their projects deposited and stored safely. Thanks to this solution, EVI-LINHD projects also improve their visibility. The specific metadata profile used in the repository is based on Dublin Core, and it is enriched with the Spanish translation of DARIAH’s Taxonomy of Digital Research Activities in the Humanities.Publicación On the optimal selection of Mel-Frequency Cepstral Coefficients for voice deepfake detection(Wiley, 2026-03-24) Falcón López, Sergio A.; Tobarra Abad, María de los Llanos; Robles Gómez, Antonio; Pastor Vargas, RafaelThe continuous evolution of techniques for generating manipulated audio, known as voice deepfakes, and the widespread availability of tools that produce convincing forgeries have created an urgent need for reliable detection methods. This work considers the dimensionality of Mel-Frequency Cepstral Coefficients (MFCCs) as a core design variable for practical, deployable systems. The aim is to identify the smallest number of coefficients that preserves detection performance across heterogeneous models while reducing computational cost, a critical factor for mobile and edge deployment. This study evaluates a hybrid setting on the ASVspoof 2019 Logical Access dataset, in which the same feature family serves as input to five traditional machine learning algorithms (Random Forest, k-Nearest Neighbors, Linear Support Vector Classification, Extreme Gradient Boosting and Support Vector Machine with radial basis function kernel) and five deep learning models (Convolutional Neural Network, Recurrent Neural Network, Convolutional Recurrent Neural Network, Xception and ResNet). Results indicate that deep models reach near-peak performance with a small number of coefficients, whereas classical methods require a larger number to achieve stable performance (except Linear Support Vector Classification, which consistently underperforms). Accordingly, 32 coefficients are considered an effective operating point for hybrid deployments. Overall, the results provide evidence to guide the selection of the number of MFCC coefficients in voice deepfake detection, aiming for efficient, reproducible and explainable systems.Publicación A WoT Platform for Supporting Full-Cycle IoT Solutions from Edge to Cloud Infrastructures: A Practical Case(MDPI, 2020-07-05) Pastor Vargas, Rafael; Tobarra Abad, María de los Llanos; Robles Gómez, Antonio; Martín Gutiérrez, Sergio; Hernández Berlinches, Roberto; Cano, Jesús; MDPI; https://orcid.org/0000-0001-6926-1311Internet of Things (IoT) learning involves the acquisition of transversal skills ranging from the development based on IoT devices and sensors (edge computing) to the connection of the devices themselves to management environments that allow the storage and processing (cloud computing) of data generated by sensors. The usual development cycle for IoT applications consists of the following three stages: stage 1 corresponds to the description of the devices and basic interaction with sensors. In stage 2, data acquired by the devices/sensors are employed by communication models from the origin edge to the management middleware in the cloud. Finally, stage 3 focuses on processing and presentation models. These models present the most relevant indicators for IoT devices and sensors. Students must acquire all the necessary skills and abilities to understand and develop these types of applications, so lecturers need an infrastructure to enable the learning of development of full IoT applications. AWeb of Things (WoT) platform named Labs of Things at UNED (LoT@UNED) has been used for this goal. This paper shows the fundamentals and features of this infrastructure, and how the different phases of the full development cycle of solutions in IoT environments are implemented using LoT@UNED. The proposed system has been tested in several computer science subjects. Students can perform remote experimentation with a collaborativeWoT learning environment in the cloud, including the possibility to analyze the generated data by IoT sensors.Publicación Distributed Parallel Hyperspectral Unmixing for Large-Scale Data in Spark Environments via Geometric Distance(Institute of Electrical and Electronics Engineers, 2025-10-23) Canada, Carlos; Paoletti, Mercedes E.; García-Flores, María B.; Tao, Xuanwen; Pastor Vargas, Rafael; Juan M. Haut; Agencia Estatal de Investigación. EspañaHyperspectral unmixing addresses the challenge of mixed pixels in hyperspectral images by identifying the number of pure pixels (endmembers), extracting their spectral signatures, and estimating their proportions (abundances) in each pixel composing the scene. Traditional hyperspectral unmixing methods often struggle with scalability and computational efficiency when dealing with gigabyte-scale datasets. In this article, we propose a distributed parallel geometric distance (DPGD) method for hyperspectral unmixing, exploring the computational power and benefits of distributed parallel processing within a distributed computing framework. The proposed DPGD leverages geometric distance measurements to accurately identify endmembers and estimate their abundances, taking into account the intrinsic similarities within hyperspectral images. This provides a clearer representation of the data structure, leading to improved unmixing accuracy. By using the Spark programming model, the computational workload is efficiently distributed across multiple nodes, significantly reducing processing time. Experimental results on real hyperspectral datasets demonstrate that DPGD scales effectively up to 32 nodes and 290.9 GB of data, achieving competitive accuracy and efficiency compared to state-of-the-art methods. The code is available at https://github.com/ccaadaro/DPDGPublicación A Cloud Game-based Educative Platform Architecture: the CyberScratch Project(MDPI, 2021) Utrilla, Alejandro; Tobarra Abad, María de los Llanos; Robles Gómez, Antonio; Pastor Vargas, Rafael; Hernández Berlinches, RobertoThe employment of modern technologies is widespread in our society, so the inclusion of practical activities for education has become essential and useful at the same time. These activities are more noticeable in Engineering, in areas such as cybersecurity, data science, artificial intelligence, etc. Additionally, these activities acquire even more relevance with a distance education methodology, as our case is. The inclusion of these practical activities has clear advantages , such as (1) promoting critical thinking and (2) improving students’ abilities and skills for their professional careers. There are several options, such as the use of remote and virtual laboratories, virtual reality and gamebased platforms, among others. This work addresses the development of a new cloud game-based educational platform, which defines a modular and flexible architecture (using light containers). This architecture provides interactive and monitoring services and data storage in a transparent way. The platform uses gamification to integrate the game as part of the instructional process. The CyberScratch project is a particular implementation of this architecture focused on cybersecurity game-based activities. The data privacy management is a critical issue for these kinds of platforms, so the architecture is designed with this feature integrated in the platform components. To achieve this goal, we first focus on all the privacy aspects for the data generated by our cloud game-based platform, by considering the European legal context for data privacy following GDPR and ISO/IEC TR 20748-1:2016 recommendations for Learning Analytics (LA). Our second objective is to provide implementation guidelines for efficient data privacy management for our cloud game-based educative platform. All these contributions are not found in current related works. The CyberScratch project, which was approved by UNED for the year 2020, considers using the xAPI standard for data handling and services for the game editor, game engine and game monitor modules of CyberScratch. Therefore, apart from considering GDPR privacy and LA recommendations, our cloud game-based architecture covers all phases from game creation to the final users’ interactions with the game.Publicación eXplicability AI (XAI) for Attack Detection toward Smart Rural Applications(Elsevier, 2025-10-20) Fernández-Morales, Enrique; Tobarra Abad, María de los Llanos; Robles Gómez, Antonio; Pastor Vargas, Rafael; Hernández Berlinches, Roberto; Sarraipa, Joao; Financiado por INCIBE en el contexto del Plan de Recuperación, Transformación y Resiliencia de la Unión Europea (NextGenerationEU/PRTR)This research evaluates the performance and computational efficiency of various AI models for intrusion detection in IoT environments, with the goal of enabling future deployment in Smart Rural scenarios. Leveraging the massive NF-UQ-NIDS-v2 dataset-comprising over 76 million labeled NetFlow records across 21 traffic classes-we benchmark five models, ranging from classical machine learning algorithms to deep learning architectures, across both high-performance and low-performance execution setups. The analysis covers standard classification metrics (accuracy, precision, recall, F1-score) and detailed resource usage indicators, including inference time, memory footprint, CPU cycles, and energy consumption per batch. Additionally, explainable AI techniques (SHAP and LIME) are employed to investigate model behavior and feature relevance under real-world constraints. Results show that classical models, particularly Random Forest and Decision Tree, achieve top-tier detection accuracy while maintaining minimal computational demands, making them strong candidates for constrained deployments. Deep learning models deliver comparable predictive performance but incur significantly higher resource consumption, requiring further optimization for practical use. Overall, this work provides a comprehensive evaluation framework and practical insights for selecting efficient and interpretable AI-based intrusion detection systems for rural and low-resource infrastructures.Publicación Smart Contracts for Managing the Chain-of-Custody of Digital Evidence: A Practical Case of Study(MDPI, 2023) Santamaría, Pablo; Tobarra Abad, María de los Llanos; Pastor Vargas, Rafael; Robles Gómez, AntonioThe digital revolution is renewing many aspects of our lives, which is also a challenge in judicial processes, such as the Chain-of-Custody (CoC) process of any electronic evidence. A CoC management system must be designed to guarantee them to maintain its integrity in court. This issue is essential for digital evidence’s admissibility and probative value. This work has built and validated a real prototype to manage the CoC process of any digital evidence. Our technological solution follows a process model that separates the evidence registry and any evidence itself for scalability purposes. It includes the development of an open-source smart contract under Quorum, a version of Ethereum oriented to private business environments. The significant findings of our analysis have been: (1) Blockchain networks can become a solution, where integrity, privacy and traceability must be guaranteed between untrustworthy parties; and (2) the necessity of promoting the standardization of CoC smart contracts with a secure, simple process logic. Consequently, these contracts should be deployed in consortium environments, where reliable, independent third parties validate the transactions without having to know their content.Publicación A Sustainable Educational Tool for Engineering Education Based on Learning Styles, AI, and Neural Networks Aligning with the UN 2030 Agenda for Sustainable Development(MDPI, 2024-10-15) Isaza Domínguez, Lauren Genith; Velasquez Clavijo, Fabian; Robles Gómez, Antonio; Pastor Vargas, RafaelThis study addresses the United Nations 2030 Agenda Sustainable Development Goals 4, 8, 10, and 12 by developing a resource-efficient tool that promotes equitable quality education and lifelong learning opportunities, supports decent work and economic growth, reduces inequalities, and ensures sustainable consumption and production patterns. This study contributes to sustainable education by providing a tool that is designed to be easy to use, easy to modify, and resource-efficient, making it accessible to institutions with limited technological resources. The tool uses artificial intelligence and a long short-term memory (LSTM) neural network to provide personalized teaching, adapting to the unique learning styles of its users. A custom survey adapted from the Felder–Silverman model was used to track weekly learning style transitions among 72 engineering students at the Faculty of Engineering at the University of Los Llanos. These data were used to build the LSTM model to predict learning style transitions over a 16-week semester. Two interfaces were created: one for instructors, integrating the LSTM model, and one for students, incorporating a custom survey. An OpenAI API-powered chat was also built into both interfaces to provide study advice to students according to their styles and enable professors to personalize their teaching methodologies in engineering education.Publicación Forensic Technologies to Automate the Acquisition of Digital Evidences(IEEE, 2022) Tobarra Abad, María de los Llanos; Robles Gómez, Antonio; Pastor Vargas, Rafael; García, DavidThe main goal of this work is to propose the automatic acquisition of evidences in a remote way. This automated capacity becomes interesting for companies with extensive networks and/or several locations, as it allows them to delegate and centralize the acquisition task at a single point in their structure, while saving time and travel costs. This research has been carried out through the initial implementation of a virtual laboratory made up of a network and different scenarios, by including an experimentation process. The virtual network includes both the machine from which automatic acquisitions are performed and the devices from retrieving the evidence. The group of devices will be made up of various experiments. The aim is to analyze the viability of the acquisition in different scenarios, since distributed networks are not homogeneous in the real worldPublicación Vulnerability Assessment of Learning Management Systems(Institute of Electrical and Electronics Engineers (IEEE), 2023-03-21) Sancristóbal Ruiz, Elio; Pastor Vargas, Rafael; Gil Ortego, Rosario; Meier, Russ; Saliah-Hassane, Hamadou; Castro Gil, Manuel AlonsoThis paper presents a vulnerability assessment of a class of web applications designed to serve educational coursework to online users. We describe the deployment landscape, the known vulnerabilities institutions must be aware of, and document how non-supervised deployment puts institutions at risk from cybercriminals.