https://reunir.unir.net/handle/123456789/19071 2026-03-12T22:43:13Z https://reunir.unir.net/handle/123456789/19242 Analysis of Artificial Intelligence Policies for Higher Education in Europe Stracke, Christian M.; Griffiths, Dai; Pappa, Dimitra; Bećirović, Senad; Polz, Edda; Perla, Loredana; Di Grassi, Annamaria; Massaro, Stefania; Skenduli, Marjana Prifti; Burgos, Daniel; Punzo, Veronica; Amram, Denise; Ziouvelou, Xenia; Katsamori, Dora; Gabriel, Sonja; Nahar, Nurun; Schleiss, Johannes; Hollins, Paul This paper analyses 15 AI policies for higher education from eight European countries, drawn from individual universities, from consortia of universities and from government agencies. Based on an overview of current research findings, it focuses the comparison of different aspects among the selected AI policies. The analysis distinguishes between four potential target groups, namely students, teachers, education managers and policy makers. The paper aims at contributing to the further development and improvement of AI policies for higher education through the identification of commonalities and gaps within the existing AI policies. Moreover, it calls for further and in particular evidence-based research to identify the potential and practical impact of AI in higher education and highlights the need to combine AI use in (higher) education with education about AI, often called as AI literacy. Submitted by Angela María Porras Ruiz (angela.porras@unir.net) on 2026-03-11T10:40:28Z No. of bitstreams: 1 Analysis of Artificial Intelligence Policies for Higher Education in Europe.pdf: 1640740 bytes, checksum: 8d471bab074d26edb071f5e645194f74 (MD5); Made available in DSpace on 2026-03-11T10:40:28Z (GMT). No. of bitstreams: 1 Analysis of Artificial Intelligence Policies for Higher Education in Europe.pdf: 1640740 bytes, checksum: 8d471bab074d26edb071f5e645194f74 (MD5) https://reunir.unir.net/handle/123456789/19239 Effects of a Flipped Classroom Learning System Integrated With ChatGPT on Students: a Survey From China Cheng, Jing; Mokmin, Nur Azlina Mohamed; Shen, Qi In design education, patterns and symbols representing traditional national cultures are often utilized as teaching materials. However, conventional teaching methods frequently fall short in aiding students' comprehension of these intricate symbolisms and abstract concepts, leading to reduced engagement and ineffective learning outcomes. Therefore, we aim to explore whether ChatGPT, as a powerful tool, can assist in solving this problem. Specifically, we integrate ChatGPT into a flipped classroom learning system to assess its effectiveness in enhancing students' understanding of traditional Chinese culture. This research contributes to the feasibility of integrating ChatGPT in design education, particularly in the context of Chinese culture. Additionally, it serves as an exploratory attempt to apply ChatGPT in teaching practices within the field of design. Submitted by Angela María Porras Ruiz (angela.porras@unir.net) on 2026-03-11T10:11:09Z No. of bitstreams: 1 Effects of a Flipped Classroom Learning System Integrated With ChatGPT on Students a Survey From China.pdf: 1495437 bytes, checksum: 6902398e040d33392da1e0585ae9e0f6 (MD5); Made available in DSpace on 2026-03-11T10:11:09Z (GMT). No. of bitstreams: 1 Effects of a Flipped Classroom Learning System Integrated With ChatGPT on Students a Survey From China.pdf: 1495437 bytes, checksum: 6902398e040d33392da1e0585ae9e0f6 (MD5) https://reunir.unir.net/handle/123456789/19238 Towards Promoting the Culture of Sharing: Using Blockchain and Artificial Intelligence in an Open Science Platform Denden, Mouna; Abed, Mourad Several studies in the literature have proposed the use of artificial intelligence (AI) tools to manage big data and further enhance collaboration between researchers on open science platforms, hence promoting the culture of safely sharing reliable data. Moreover, some other studies further proposed the use of blockchain technology to secure data, provide transparency in data analysis, and also keep track of all collaborations within open science platforms. Despite the importance of AI and blockchain technology in open science platforms, no study, to the best of our knowledge, has implemented and discussed the benefits of using both technologies together or how blockchain can enhance AI systems in open science. Therefore, to address this research gap, this study presents a newly developed open science platform that harnesses the power of AI and blockchain technologies to promote and foster a culture of sharing and seamless collaboration among universities worldwide. This platform was then validated through focus group analysis from the European University for Customised Education (EUNICE) partners, which is the project context of this present study. The findings revealed that the use of AI and blockchain enabled researchers and institutions to share open science more effectively. Specifically, the use of AI features in Open REUNICE enhanced data management processes, particularly by improving metadata accuracy, searchability and reusability, thereby addressing critical needs in research workflows. Additionally, the use of Blockchain was found to play a critical role in addressing legal challenges and enhancing user trust. Submitted by Angela María Porras Ruiz (angela.porras@unir.net) on 2026-03-11T10:06:33Z No. of bitstreams: 1 Towards Promoting the Culture of Sharing Using Blockchain and Artificial Intelligence in an Open Science Platform.pdf: 1109690 bytes, checksum: 6f256836fc257589793f63badfd93289 (MD5); Made available in DSpace on 2026-03-11T10:06:33Z (GMT). No. of bitstreams: 1 Towards Promoting the Culture of Sharing Using Blockchain and Artificial Intelligence in an Open Science Platform.pdf: 1109690 bytes, checksum: 6f256836fc257589793f63badfd93289 (MD5) https://reunir.unir.net/handle/123456789/19237 Gaming as a Medium for the Expression of Citizens' Views on Environmental Dilemmas. Griffiths, Dai; Ower, Jude; Hollins, Paul; Garg, Anchal The decline of traditional media and channels of communication has led to policymakers experiencing difficulty in understanding public sentiment. A case study was conducted to explore how games-based activities can be used to provide a link between citizens and policy makers. A system developed by PlanetPlay, and extended in the GREAT project, was used to embed a survey in the game SMITE. The intervention and survey questions were designed in collaboration with the United Nations Development Programme (UNDP) and the Hi-Rez game studio. The effectiveness of the infrastructure and the collaborative approach were demonstrated. The results revealed some significant differences in views on climate change between different age groups, genders, and education level. However, the data was heavily skewed towards males in the 18-35 age group, and to respondents in the United States, which limited the generalizability of the findings. It was concluded that in-game placement in collaboration with games studios is more effective than paid placement, and that a wider variety of games is needed to ensure that a study has an adequate range of respondent profiles. Finally, reflections are offered on the possible role of artificial intelligence in gathering such data. Submitted by Angela María Porras Ruiz (angela.porras@unir.net) on 2026-03-11T10:02:36Z No. of bitstreams: 1 Gaming as a Medium for the Expression of Citizens Views on Environmental Dilemmas.pdf: 557475 bytes, checksum: a751841e0440836bf9cf92a475c2c8c4 (MD5); Made available in DSpace on 2026-03-11T10:02:36Z (GMT). No. of bitstreams: 1 Gaming as a Medium for the Expression of Citizens Views on Environmental Dilemmas.pdf: 557475 bytes, checksum: a751841e0440836bf9cf92a475c2c8c4 (MD5) https://reunir.unir.net/handle/123456789/19236 Youth Expectations and Perceptions of Generative Artificial Intelligence in Higher Education Cotino Arbelo, Andrea E.; González González, Carina S.; Molina Gil, Jezabel Artificial Intelligence (AI) is not a recent innovation, what’s new is how accessible its features have become across multiple devices, apps, and services. Sensationalistic news can distort public perception by exaggerating AI’s capabilities and risks. This leads to misconceptions and unrealistic expectations, causing misunderstandings about the true nature and limitation of these tools. Such distortions can undermine trust and hinder the effective adoption and integration of AI into society. This study aims to address this issue by exploring the expectations and perceptions of young individuals regarding Generative Artificial Intelligence (GAI) tools. It explores their understanding of GAI and related devices, such as virtual assistants, chatbots, and social robots, which can incorporate GAI. A total of N=100 university students engaged in this study by completing a digital questionnaire distributed through the virtual campus of the University of La Laguna. The quantitative analysis uncovered a significant gap in participants’ understanding of GAI terminology and its underlying mechanisms. Additionally, it shed light on a noteworthy gender based discrepancy in the expressed concerns. Participants commonly recognized their ability to communicate effectively with GAI, asserting that such interactions enhance their emotional well-being. Notably, virtual assistants and chatbots were perceived as more valuable tools compared to social robots within the educational realm. Submitted by Angela María Porras Ruiz (angela.porras@unir.net) on 2026-03-11T09:58:14Z No. of bitstreams: 1 Youth Expectations and Perceptions of Generative Artificial Intelligence in Higher Education.pdf: 799846 bytes, checksum: 2e0246100663ce629ce02052db2d1438 (MD5); Made available in DSpace on 2026-03-11T09:58:14Z (GMT). No. of bitstreams: 1 Youth Expectations and Perceptions of Generative Artificial Intelligence in Higher Education.pdf: 799846 bytes, checksum: 2e0246100663ce629ce02052db2d1438 (MD5) https://reunir.unir.net/handle/123456789/19234 Sentiment Analysis With Transformers Applied to Education: Systematic Review Pilicita Garrido, Anabel; Barra, Enrique Sentiment analysis, empowered by artificial intelligence, can play a critical role in assessing the impact of cultural factors on the advancement of Open Science and artificial intelligence. Additionally, it can offer valuable insights into the open data gathered within educational contexts. This article presents a systematic review of the use of Transformers models in sentiment analysis in education. A systematic review approach was used to analyze 41 articles from recognized digital databases. The results of the review provide a comprehensive understanding of previous research related to the use of Transformers models in education for the task of sentiment analysis, their benefits, challenges, as well as future areas of research that can lay the foundation for a more sustainable and effective education system. Submitted by Angela María Porras Ruiz (angela.porras@unir.net) on 2026-03-11T09:53:53Z No. of bitstreams: 1 Sentiment Analysis With Transformers Applied to Education Systematic Review.pdf: 626731 bytes, checksum: e9a28424b977fb97e38f6e4789327970 (MD5); Made available in DSpace on 2026-03-11T09:53:53Z (GMT). No. of bitstreams: 1 Sentiment Analysis With Transformers Applied to Education Systematic Review.pdf: 626731 bytes, checksum: e9a28424b977fb97e38f6e4789327970 (MD5) https://reunir.unir.net/handle/123456789/19233 AI Hallucinations? What About Human Hallucination?! Addressing Human Imperfection Is Needed for an Ethical AI Tlili, Ahmed; Burgos, Daniel This study discusses how the human imperfection nature, also known as the human hallucination, could contribute to or emphasize technology (generally) and Artificial Intelligence (AI, particularly) hallucination. While the ongoing debate puts more efforts on improving AI for its ethical use, a shift should be made to also cover us, humans, who are the technology designer, developer, and user. Identifying and understanding the link between human and AI hallucination will ultimately help to develop effective and safe AI-powered systems that could have some positive societal impact in the long run. Submitted by Angela María Porras Ruiz (angela.porras@unir.net) on 2026-03-11T09:50:39Z No. of bitstreams: 1 AI Hallucinations What About Human Hallucination Addressing Human Imperfection Is Needed for an Ethical AI.pdf: 466298 bytes, checksum: 1bb9a9be820b29e6f6a1e9075c541f1e (MD5); Made available in DSpace on 2026-03-11T09:50:39Z (GMT). No. of bitstreams: 1 AI Hallucinations What About Human Hallucination Addressing Human Imperfection Is Needed for an Ethical AI.pdf: 466298 bytes, checksum: 1bb9a9be820b29e6f6a1e9075c541f1e (MD5) https://reunir.unir.net/handle/123456789/19232 Improved Fine-Tuned Reinforcement Learning From Human Feedback Using Prompting Methods for News Summarization Pulari, Sini Raj; Umadevi, Maramreddy; Vasudevan, Shriram K. ChatGPT uses a generative pretrained transformer neural network model, which is under the larger umbrella of generative models. One major boom after ChatGPT is the advent of prompt engineering, which is the most critical part of ChatGPT that utilizes Large Language Models (LLM) and helps ChatGPT provide the desired outputs based on the style and tone of interactions carried out with it. Reinforcement learning from human feedback (RLHF) was used as the major aspect for fine-tuning LLM-based models. This work proposes a human selection strategy that is incorporated in the RLHF process to prevent undesirable consequences of the rightful choice of human reviewers for feedback. H-Rouge is a new metric proposed for humanized AI systems. A detailed evaluation of State-of-the-art summarization algorithms and prompt-based methods have been provided as part of the article. The proposed methods have introduced a strategy for human selection of RLHF models which employs multi-objective optimization to balance various goals encountered during the process with H-Rouge. This article will help nuance readers conduct research in the field of text summarization to start with prompt engineering in the summarization field, and future work will help them proceed in the right direction of research. Submitted by Angela María Porras Ruiz (angela.porras@unir.net) on 2026-03-11T09:47:13Z No. of bitstreams: 1 Improved Fine-Tuned Reinforcement Learning From Human Feedback Using Prompting Methods for News Summarization.pdf: 1439406 bytes, checksum: edccf242c6118f5570384935df8cc233 (MD5); Made available in DSpace on 2026-03-11T09:47:13Z (GMT). No. of bitstreams: 1 Improved Fine-Tuned Reinforcement Learning From Human Feedback Using Prompting Methods for News Summarization.pdf: 1439406 bytes, checksum: edccf242c6118f5570384935df8cc233 (MD5) https://reunir.unir.net/handle/123456789/19231 Aligning Figurative Paintings With Their Sources for Semantic Interpretation Aslan, Sinem; Steels, Luc This paper reports steps in probing the artistic methods of figurative painters through computational algorithms. We explore a comparative method that investigates the relation between the source of a painting, typically a photograph or an earlier painting, and the painting itself. A first crucial step in this process is to find the source and to crop, standardize and align it to the painting so that a comparison becomes possible. The next step is to apply different low-level algorithms to construct difference maps for color, edges, texture, brightness, etc. From this basis, various subsequent operations become possible to detect and compare features of the image, such as facial action units and the emotions they signify. This paper demonstrates a pipeline we have built and tested using paintings by a renowned contemporary painter Luc Tuymans. We focus in this paper particularly on the alignment process, on edge difference maps, and on the utility of the comparative method for bringing out the semantic significance of a painting. Submitted by Angela María Porras Ruiz (angela.porras@unir.net) on 2026-03-11T09:43:30Z No. of bitstreams: 1 Aligning Figurative Paintings With Their Sources for Semantic Interpretation.pdf: 5986227 bytes, checksum: ed032401a3e4dfbcbb8bae7ff42d9e37 (MD5); Made available in DSpace on 2026-03-11T09:43:30Z (GMT). No. of bitstreams: 1 Aligning Figurative Paintings With Their Sources for Semantic Interpretation.pdf: 5986227 bytes, checksum: ed032401a3e4dfbcbb8bae7ff42d9e37 (MD5) https://reunir.unir.net/handle/123456789/19229 Spatial-Aware Multi-Level Parsing Network for Human-Object Interaction Su, Zhan; Yu, Ruiyun; Zou, Shihao; Guo, Bingyang; Cheng, Li Human-Object Interaction (HOI) detection focuses on human-centered visual relationship detection, which is a challenging task due to the complexity and diversity of image content. Unlike most recent HOI detection works that only rely on paired instance-level information in the union range, our proposed Spatial-aware Multilevel Parsing Network (SMPNet) uses a multi-level information detection strategy, including instance-level visual features of detected human-object pair, part-level related features of the human body, and scene-level features extracted by the graph neural network. After fusing the three levels of features, the HOI relationship is predicted. We validate our method on two public datasets, V-COCO and HICO-DET. Compared with prior works, our proposed method achieves the state-of-the-art results on both datasets in terms of mAProle, which demonstrates the effectiveness of our proposed multi-level information detection strategy Submitted by Angela María Porras Ruiz (angela.porras@unir.net) on 2026-03-11T09:40:02Z No. of bitstreams: 1 Spatial-Aware Multi-Level Parsing Network for Human-Object Interaction.pdf: 8072831 bytes, checksum: f382c692472bfba9537e7a5ab5a4167d (MD5); Made available in DSpace on 2026-03-11T09:40:02Z (GMT). No. of bitstreams: 1 Spatial-Aware Multi-Level Parsing Network for Human-Object Interaction.pdf: 8072831 bytes, checksum: f382c692472bfba9537e7a5ab5a4167d (MD5) https://reunir.unir.net/handle/123456789/19228 A Hybrid Multi-Person Fall Detection Scheme Based on Optimized YOLO and ST-GCN Liu, Lei; Sun, Yeguo; Ge, Xianlei Human falls are a serious health issue for elderly and disabled people living alone. Studies have shown that if fallers could be helped immediately after a fall, it would greatly reduce their risk of death and the percentage of them requiring long-term treatment. As a real-time automatic fall detection solution, vision-based human fall detection technology has received extensive attention from researchers. In this paper, a hybrid model based on YOLO and ST-GCN is proposed for multi-person fall detection application scenarios. The solution uses the ST-GCN model based on a graph convolutional network to detect the fall action, and enhances the model with YOLO for accurate and fast recognition of multi-person targets. Meanwhile, our scheme accelerates the model through optimization methods to meet the model's demand for lightweight and real-time performance. Finally, we conducted performance tests on the designed prototype system and using both publicly available single-person datasets and our own multi-person dataset. The experimental results show that under better environmental conditions, our model possesses high detection accuracy compared to state-of-the-art schemes, while it significantly outperforms other models in terms of inference speed. Therefore, this hybrid model based on YOLO and ST-GCN, as a preliminary attempt, provides a new solution idea for multi-person fall detection for the elderly. Submitted by Angela María Porras Ruiz (angela.porras@unir.net) on 2026-03-11T09:36:18Z No. of bitstreams: 1 The Application of Deep Learning for Classification of Alzheimer's Disease Stages by Magnetic Resonance Imaging Data.pdf: 2311411 bytes, checksum: 0274fd67e29a8306be9a353c3585298e (MD5); Made available in DSpace on 2026-03-11T09:36:18Z (GMT). No. of bitstreams: 1 The Application of Deep Learning for Classification of Alzheimer's Disease Stages by Magnetic Resonance Imaging Data.pdf: 2311411 bytes, checksum: 0274fd67e29a8306be9a353c3585298e (MD5) https://reunir.unir.net/handle/123456789/19227 The Application of Deep Learning for Classification of Alzheimer's Disease Stages by Magnetic Resonance Imaging Data Irfan, Muhammad; Shahrestani, Seyed; ElKhodr, Mahmoud Detecting Alzheimer’s disease (AD) in its early stages is essential for effective management, and screening for Mild Cognitive Impairment (MCI) is common practice. Among many deep learning techniques applied to assess brain structural changes, Magnetic Resonance Imaging (MRI) and Convolutional Neural Networks (CNN) have grabbed research attention because of their excellent efficiency in automated feature learning of a variety of multilayer perceptron. In this study, various CNNs are trained to predict AD on three different views of MRI images, including Sagittal, Transverse, and Coronal views. This research use T1-Weighted MRI data of 3 years composed of 2182 NIFTI files. Each NIFTI file presents a single patient's Sagittal, Transverse, and Coronal views. T1-Weighted MRI images from the ADNI database are first preprocessed to achieve better representation. After MRI preprocessing, large slice numbers require a substantial computational cost during CNN training. To reduce the slice numbers for each view, this research proposes an intelligent probabilistic approach to select slice numbers such that the total computational cost per MRI is minimized. With hyperparameter tuning, batch normalization, and intelligent slice selection and cropping, an accuracy of 90.05% achieve with the Transverse, 82.4% with Sagittal, and 78.5% with Coronal view, respectively. Moreover, the views are stacked together and an accuracy of 92.21% is achived for the combined views. In addition, results are compared with other studies to show the performance of the proposed approach for AD detection. Submitted by Angela María Porras Ruiz (angela.porras@unir.net) on 2026-03-11T09:32:21Z No. of bitstreams: 1 The Application of Deep Learning for Classification of Alzheimer's Disease Stages by Magnetic Resonance Imaging Data.pdf: 2311411 bytes, checksum: 0274fd67e29a8306be9a353c3585298e (MD5); Made available in DSpace on 2026-03-11T09:32:21Z (GMT). No. of bitstreams: 1 The Application of Deep Learning for Classification of Alzheimer's Disease Stages by Magnetic Resonance Imaging Data.pdf: 2311411 bytes, checksum: 0274fd67e29a8306be9a353c3585298e (MD5) https://reunir.unir.net/handle/123456789/19226 A Smart Healthcare System Using Consumer Electronics and Federated Learning to Automatically Diagnose Diabetic Foot Ulcers Das, Sujit Kumar; Moparthi, Nageswara Rao; Namasudra, Suyel; González Crespo, Rubén; Taniar, David Privacy breaches on sensitive and widely distributed health data in consumer electronics (CE) demand novel strategies to protect privacy with correctness and proper operation maintenance. This work presents a scalable Federated Learning (FL) framework-based smart healthcare approach. Remote medical facilities frequently struggle with imbalanced datasets, including intermittent client connections to the FL global server. The proposed approach handled intermittent clients with diabetic foot ulcers (DFU) images. A data augmentation approach proposes to handle class imbalance problems during local model training. Also, a novel Convolutional Neural Network (CNN) architecture, ResKNet (K=4), is designed for client-side model training. The ResKNet is a sequence of distinctive residual blocks with 2D convolution, batch normalization, LeakyReLU activation, and skip connections (convolutional and identity). The proposed approach is evaluated for various client counts (5,10,15, and 20) and multiple test dataset sizes. The proposed framework can leverage consumer electronic devices and ensure secure data sharing among multiple sources. The potential of integrating the proposed approach with smartphones and wearable devices to provide highly secure data transmission is very high. The approach also helps medical institutions collaborate and develop a robust patient diagnostic model. Submitted by Angela María Porras Ruiz (angela.porras@unir.net) on 2026-03-11T09:28:05Z No. of bitstreams: 1 A Smart Healthcare System Using Consumer Electronics and Federated Learning to Automatically Diagnose Diabetic Foot Ulcers.pdf: 3506674 bytes, checksum: 08199520417a20c2f70a03c4565bf64f (MD5); Made available in DSpace on 2026-03-11T09:28:05Z (GMT). No. of bitstreams: 1 A Smart Healthcare System Using Consumer Electronics and Federated Learning to Automatically Diagnose Diabetic Foot Ulcers.pdf: 3506674 bytes, checksum: 08199520417a20c2f70a03c4565bf64f (MD5) https://reunir.unir.net/handle/123456789/19225 A Smart Healthcare System Using Consumer Electronics and Federated Learning to Automatically Diagnose Diabetic Foot Ulcers Das, Sujit Kumar; Moparthi, Nageswara Rao; Namasudra, Suyel; González Crespo, Rubén; Taniar, David Privacy breaches on sensitive and widely distributed health data in consumer electronics (CE) demand novel strategies to protect privacy with correctness and proper operation maintenance. This work presents a scalable Federated Learning (FL) framework-based smart healthcare approach. Remote medical facilities frequently struggle with imbalanced datasets, including intermittent client connections to the FL global server. The proposed approach handled intermittent clients with diabetic foot ulcers (DFU) images. A data augmentation approach proposes to handle class imbalance problems during local model training. Also, a novel Convolutional Neural Network (CNN) architecture, ResKNet (K=4), is designed for client-side model training. The ResKNet is a sequence of distinctive residual blocks with 2D convolution, batch normalization, LeakyReLU activation, and skip connections (convolutional and identity). The proposed approach is evaluated for various client counts (5,10,15, and 20) and multiple test dataset sizes. The proposed framework can leverage consumer electronic devices and ensure secure data sharing among multiple sources. The potential of integrating the proposed approach with smartphones and wearable devices to provide highly secure data transmission is very high. The approach also helps medical institutions collaborate and develop a robust patient diagnostic model. Submitted by Angela María Porras Ruiz (angela.porras@unir.net) on 2026-03-11T09:22:15Z No. of bitstreams: 1 A Smart Healthcare System Using Consumer Electronics and Federated Learning to Automatically Diagnose Diabetic Foot Ulcers.pdf: 3506674 bytes, checksum: 08199520417a20c2f70a03c4565bf64f (MD5); Made available in DSpace on 2026-03-11T09:22:15Z (GMT). No. of bitstreams: 1 A Smart Healthcare System Using Consumer Electronics and Federated Learning to Automatically Diagnose Diabetic Foot Ulcers.pdf: 3506674 bytes, checksum: 08199520417a20c2f70a03c4565bf64f (MD5) https://reunir.unir.net/handle/123456789/19224 Editor’s Note Verdú, Elena The International Journal of Interactive Multimedia and Artificial Intelligence (IJIMAI) is a diamond open access journal which provides an interdisciplinary forum in which scientists and professionals can share their research results and report new advances on artificial intelligence tools, theory, methodologies, systems, architectures integrating multiple technologies, problems including demonstrations of effectiveness, or tools that use AI with interactive multimedia techniques. The journal is supported by Universidad Internacional de La Rioja (UNIR) and by all those members of this multicultural community who, with a sense of commitment to the development of science, dedicate their knowledge and time to authoring, editing and reviewing tasks, and without whom this knowledge sharing project would not be possible. This regular issue begins with a series of five articles covering key advancements in the area of computing vision. In the following article, we move from the area of computer vision to another fast developing area, which is natural language processing (NLP). The following articles correspond to a monograph section on the Effects of Culture on Open Science and Artificial Intelligence in Education, compiled and edited by Tlili, Burgos and Kinshuk. Submitted by Angela María Porras Ruiz (angela.porras@unir.net) on 2026-03-11T09:12:36Z No. of bitstreams: 1 Editor’s Note.pdf: 59199 bytes, checksum: cf0c05a5b35d1993ade643aa8bf7eeb8 (MD5); Made available in DSpace on 2026-03-11T09:12:36Z (GMT). No. of bitstreams: 1 Editor’s Note.pdf: 59199 bytes, checksum: cf0c05a5b35d1993ade643aa8bf7eeb8 (MD5) https://reunir.unir.net/handle/123456789/19222 Evaluating Customer Segmentation Techniques in the Retail Sector Diyabi, Nur; Çakır, Duygu; Gül, Ömer Melih; Aytekin, Tevfik; Kadry, Seifedine In the current competitive corporate landscape, understanding client preferences and adapting marketing strategies accordingly has become crucial. This study evaluates the effectiveness of four machine learning algorithms (K-Means, Density-Based Spatial Clustering of Applications with Noise (DBSCAN), Gaussian Mixture Models (GMM), and Self-Organizing Maps (SOM)) for customer segmentation in the Turkish retail market. Two datasets were analyzed: a large-scale Turkish market sales dataset and a focused marketing campaign dataset. The research employed a comprehensive methodology encompassing data preparation, algorithm application, and performance evaluation using metrics such as the Calinski-Harabasz Index and Davies- Bouldin score. Results indicate that K-Means demonstrated superior performance in terms of interpretability and statistical validity. DBSCAN showed strengths in identifying non-spherical clusters, while GMM and SOM provided more granular segmentation. The findings offer actionable insights for Turkish retailers to optimize marketing strategies and enhance customer relationship management. This study contributes to the field of retail analytics by providing a methodological framework for evaluating customer segmentation techniques in specific market contexts. Submitted by Angela María Porras Ruiz (angela.porras@unir.net) on 2026-03-11T08:30:19Z No. of bitstreams: 1 Evaluating Customer Segmentation Techniques in the Retail Sector.pdf: 1344269 bytes, checksum: 24a95aafc7d1bc90e48f00971b2a85f4 (MD5); Made available in DSpace on 2026-03-11T08:30:19Z (GMT). No. of bitstreams: 1 Evaluating Customer Segmentation Techniques in the Retail Sector.pdf: 1344269 bytes, checksum: 24a95aafc7d1bc90e48f00971b2a85f4 (MD5) https://reunir.unir.net/handle/123456789/19221 A Multi-Session Evaluation of a Haptic Device in Normal and Critical Conditions: a Mars Analog Mission Manon, Julie; Vanderdonckt, Jean; Saint Guillain, Michael; Pletser, Vladimir; Wain, Cyril; Jacobs, Jean; Comein, Audrey; Drouet, Sirga; Meert, Julien; Sanchez Casla, Ignacio; Cartiaux, Olivier; Cornu, Olivier While visual interaction is typically evaluated as an instantaneous, one-shot activity that considers only a snapshot of factors, haptic interaction is more challenging to evaluate as it involves a continuous touch process evolving over time. To better understand how to evaluate haptic interaction, this paper performs a multisession evaluation of a haptic device to be used by astronauts in future lunar and Mars missions, based on eight factors. Three groups of two members (???? = 6 ) applied, either as operator or assistant, a newly developed external fixator (EZExFix) to fix a fracture of the tibial shaft. Astronauts had different levels of expertise, i.e., in anatomy, mechanical engineering, and without, and participated in eight timed runs. Among these eight matches, four sessions were conducted with different time frames and compared to a stress test, a reproduction of the experiment in very stressful conditions, and a session simulating critical conditions in an extra-vehicular activity. Submitted by Angela María Porras Ruiz (angela.porras@unir.net) on 2026-03-11T08:25:12Z No. of bitstreams: 1 A Multi-Session Evaluation of a Haptic Device in Normal and Critical Conditions.pdf: 877083 bytes, checksum: 85bc2ce8bad2a25e96029789a9e17eec (MD5); Made available in DSpace on 2026-03-11T08:25:12Z (GMT). No. of bitstreams: 1 A Multi-Session Evaluation of a Haptic Device in Normal and Critical Conditions.pdf: 877083 bytes, checksum: 85bc2ce8bad2a25e96029789a9e17eec (MD5) https://reunir.unir.net/handle/123456789/19220 Reliability of IBM’s Public Quantum Computers Pérez Antón, Raquel; Corbi, Alberto; López Sánchez, José Ignacio; Burgos, Daniel One of the challenges of the current ecosystem of quantum computers (QC) is the stabilization of the coherence associated with the entanglement of the states of their inner qubits. In this empirical study, we monitor the reliability of IBM’s public-access QCs network on a daily basis. Each of these state-of-the-art machines has a totally different qubit association, and this entails that for a given (same) input program, they may output a different set of probabilities for the assembly of results (including both the right and the wrong ones). Although we focus on the computing structure provided by the “Big Blue” company, our survey can be easily transferred to other currently available quantum mainframes. In more detail, we probe these quantum processors with an ad hoc designed computationally demanding quaternary search algorithm. As stated, this quantum program is executed every 24 hours (for nearly 100 days) and its goal is to put to the limit the operational capacity of this novel and genuine type of equipment. Next, we perform a comparative analysis of the obtained results according to the singularities of each computer and over the total number of executions. In addition, we subsequently apply (for 50 days) an improvement filtering to perform noise mitigation on the results obtained proposed by IBM. The Yorktown 5-qubit computer reaches noise filtering of up to 33% in one day, that is, a 90% confidence level is reached in the expected results. From our continuous and long-term tests, we derive that room still exists regarding the improvement of quantum calculators in order to guarantee enough confidence in the returned outcomes. Submitted by Angela María Porras Ruiz (angela.porras@unir.net) on 2026-03-11T08:16:47Z No. of bitstreams: 1 Reversible Image Watermarking Using Modified Quadratic Difference Expansion and Hybrid Optimization Technique.pdf: 1980156 bytes, checksum: 0aae57c42bf1ac2d4fbc69ed5fd20920 (MD5); Made available in DSpace on 2026-03-11T08:16:47Z (GMT). No. of bitstreams: 1 Reversible Image Watermarking Using Modified Quadratic Difference Expansion and Hybrid Optimization Technique.pdf: 1980156 bytes, checksum: 0aae57c42bf1ac2d4fbc69ed5fd20920 (MD5) https://reunir.unir.net/handle/123456789/19219 Reversible Image Watermarking Using Modified Quadratic Difference Expansion and Hybrid Optimization Technique Lakshmi, H. R.; Borra, Surekha With increasing copyright violation cases, watermarking of digital images is a very popular solution for securing online media content. Since some sensitive applications require image recovery after watermark extraction, reversible watermarking is widely preferred. This article introduces a Modified Quadratic Difference Expansion (MQDE) and fractal encryption-based reversible watermarking for securing the copyrights of images. First, fractal encryption is applied to watermarks using Tromino's L-shaped theorem to improve security. In addition, Cuckoo Search-Grey Wolf Optimization (CSGWO) is enforced on the cover image to optimize block allocation for inserting an encrypted watermark such that it greatly increases its invisibility. While the developed MQDE technique helps to improve coverage and visual quality, the novel data-driven distortion control unit ensures optimal performance. The suggested approach provides the highest level of protection when retrieving the secret image and original cover image without losing the essential information, apart from improving transparency and capacity without much tradeoff. The simulation results of this approach are superior to existing methods in terms of embedding capacity. With an average PSNR of 67 dB, the method shows good imperceptibility in comparison to other schemes. Submitted by Angela María Porras Ruiz (angela.porras@unir.net) on 2026-03-11T08:12:00Z No. of bitstreams: 1 Reversible Image Watermarking Using Modified Quadratic Difference Expansion and Hybrid Optimization Technique.pdf: 1980156 bytes, checksum: 0aae57c42bf1ac2d4fbc69ed5fd20920 (MD5); Made available in DSpace on 2026-03-11T08:12:00Z (GMT). No. of bitstreams: 1 Reversible Image Watermarking Using Modified Quadratic Difference Expansion and Hybrid Optimization Technique.pdf: 1980156 bytes, checksum: 0aae57c42bf1ac2d4fbc69ed5fd20920 (MD5) https://reunir.unir.net/handle/123456789/19218 Simulations for the Precise Modeling of Exercises Including Time, Grades and Number of Attempts Jiménez-Macías, Alberto; Muñoz-Merino, Pedro J.; Delgado Kloos, Carlos Students’ interactions with exercises can reveal interesting features that can be used to redesign or effectively use the exercises during the learning process. The precise modeling of exercises includes how grades can evolve, depending on the number of attempts and time spent on the exercises. A missing aspect is how a precise relationship among grades, number of attempts, and time spent can be inferred from student interactions with exercises using machine learning methods, and how it differs depending on different factors. In this study, we analyzed the application of different machine-learning methods for modeling different scenarios by varying the probability of answering correctly, dataset sizes, and distributions. The results show that the model converged when the probability of random guessing was low. For exercises with an average of 2 attempts, the model converged to 200 interactions. However, increasing the number of interactions beyond 200 does not affect the accuracy of the model. Submitted by Angela María Porras Ruiz (angela.porras@unir.net) on 2026-03-11T08:08:17Z No. of bitstreams: 1 Simulations for the Precise Modeling of Exercises Including Time, Grades and Number of Attempts.pdf: 730266 bytes, checksum: c4f311227cef1c69af7b2270a0dabecc (MD5); Made available in DSpace on 2026-03-11T08:08:17Z (GMT). No. of bitstreams: 1 Simulations for the Precise Modeling of Exercises Including Time, Grades and Number of Attempts.pdf: 730266 bytes, checksum: c4f311227cef1c69af7b2270a0dabecc (MD5) https://reunir.unir.net/handle/123456789/19217 Learning Analytics Icons: Easy Comprehension of Data Treatment Amo-Filva, Daniel; Alier, Marc; Fonseca, David; Garcia-Peñalvo, Francisco José; Casañ, María José The Learning Analytics approach adopted in education implies the gathering and processing of sensitive information and the generation of student profiles, which may have direct or indirect dire consequences for the students. The Educational institutions must manage this data processing according to the General Data Protection Regulation, respecting its principles of fairness when it comes to information gathering and processing. This implies that the students must be well informed and give explicit consent before their information is gathered and processed. The GDPR propose the usage of recognizable standardized icons to facilitate a general understanding and awareness of how personal data is deemed to be processed in each application context, like an online course. This paper presents a project that aims to provide a set of icons to inform about the treatment of educational data in the Learning Analytics processes and a survey about the student's comprehension of the icons, their meaning, and implications for their privacy and confidentiality. The result presented is a set of icons ready to be integrated into educational environments that apply Learning Analytics to increase transparency and facilitate the understanding of data processing. Submitted by Angela María Porras Ruiz (angela.porras@unir.net) on 2026-03-11T08:00:02Z No. of bitstreams: 1 Learning Analytics Icons Easy Comprehension of Data Treatment.pdf: 742924 bytes, checksum: d7cf786ac3a600774d92a70ed26dd70f (MD5); Made available in DSpace on 2026-03-11T08:00:02Z (GMT). No. of bitstreams: 1 Learning Analytics Icons Easy Comprehension of Data Treatment.pdf: 742924 bytes, checksum: d7cf786ac3a600774d92a70ed26dd70f (MD5) https://reunir.unir.net/handle/123456789/19215 An Effective Prediction Approach for the Management of Children Victims of Road Accidents Saadi, F.; Atmani, B.; Henni, F.; Benfriha, H.; Addou, Z.; Guerbouz, R. Road traffic generates a considerable number of accidents each year. The management of injuries caused by these accidents is becoming a real public health problem. Faced with this latter, we propose a new clinical decision making approach based on case-based reasoning (CBR) and data mining (DM) techniques to speed up and improve the care of an injured child. The main idea is to preprocess the dataset before using K Nearest Neighbor (KNN) Classification Model. In this paper, an efficient predictive model is developed to predict the admission procedure of a child victim of a traffic accident in pediatric intensive care units. The evaluation of the proposed model is conducted on a real dataset elaborated by the authors and validated by statistical analysis. This novel model executes a selection of relevant attributes using data mining technique and integrates a CBR system to retrieve similar cases from an archive of cases of patients successfully treated with the proposed treatment plan. The results revealed that the proposed approach outperformed other models and the results of previous studies by achieving an accuracy of 91.66%. Submitted by Angela María Porras Ruiz (angela.porras@unir.net) on 2026-03-10T16:52:58Z No. of bitstreams: 1 An Effective Prediction Approach for the Management of Children Victims of Road Accidents.pdf: 456184 bytes, checksum: 3965144c96757b44a1f1e1382039e0ad (MD5); Made available in DSpace on 2026-03-10T16:52:58Z (GMT). No. of bitstreams: 1 An Effective Prediction Approach for the Management of Children Victims of Road Accidents.pdf: 456184 bytes, checksum: 3965144c96757b44a1f1e1382039e0ad (MD5) https://reunir.unir.net/handle/123456789/19214 Traffic Optimization Through Waiting Prediction and Evolutive Algorithms García, Francisco Traffic optimization systems require optimization procedures to optimize traffic light timing settings in order to improve pedestrian and vehicle mobility. Traffic simulators allow obtaining accurate estimates of traffic behavior by applying different timing configurations, but require considerable computational time to perform validation tests. For this reason, this project proposes the development of traffic optimizations based on the estimation of vehicle waiting times through the use of different prediction techniques and the use of this estimation to subsequently apply evolutionary algorithms that allow the optimizations to be carried out. The combination of these two techniques leads to a considerable reduction in calculation time, which makes it possible to apply this system at runtime. The tests have been carried out on a real traffic junction on which different traffic volumes have been applied to analyze the performance of the system. Submitted by Angela María Porras Ruiz (angela.porras@unir.net) on 2026-03-10T16:44:42Z No. of bitstreams: 1 Traffic Optimization Through Waiting Prediction and Evolutive Algorithms.pdf: 711723 bytes, checksum: c097d0a6f750ca2a64f026156e4c9f85 (MD5); Made available in DSpace on 2026-03-10T16:44:42Z (GMT). No. of bitstreams: 1 Traffic Optimization Through Waiting Prediction and Evolutive Algorithms.pdf: 711723 bytes, checksum: c097d0a6f750ca2a64f026156e4c9f85 (MD5) https://reunir.unir.net/handle/123456789/19213 Use of Data Mining for Intelligent Evaluation of Imputation Methods la Red Martínez, David L.; Primorac, Carlos R. In real-world situations, researchers frequently face the difficulty of missing values (MV), i.e., values not observed in a data set. Data imputation techniques allow the estimation of MV using different algorithms, by means of which important data can be imputed for a particular instance. Most of the literature in this field deals with different imputation methods. However, few studies deal with a comparative evaluation of the different methods as to provide more appropriate guidelines for the selection of the method to be applied to impute data for specific situations. The objective of this work is to show a methodology for evaluating the performance of imputation methods by means of new metrics derived from data mining processes, using quality metrics of data mining models. We started from the complete dataset that was amputated with different amputation mechanisms to generate 63 datasets with MV; these were imputed using Median, k-NN, k-Means and Hot-Deck imputation methods. The performance of the imputation methods was evaluated using new metrics derived from quality metrics of the data mining processes, performed with the original full file and with the imputed files. This evaluation is not based on measuring the error when imputing (usual operation), but on considering the similarity of the values of the quality metrics of the data mining processes obtained with the original file and with the imputed files. The results show that –globally considered and according to the new proposed metric, the imputation methods that showed the best performance were k-NN and k-Means. An additional advantage of the proposed methodology is that it provides predictive data mining models that can be used a posteriori. Submitted by Angela María Porras Ruiz (angela.porras@unir.net) on 2026-03-10T16:41:16Z No. of bitstreams: 1 Use of Data Mining for Intelligent Evaluation of Imputation Methods.pdf: 791527 bytes, checksum: 3cd9625f80fe5307b0d325d9a030d7a7 (MD5); Made available in DSpace on 2026-03-10T16:41:16Z (GMT). No. of bitstreams: 1 Use of Data Mining for Intelligent Evaluation of Imputation Methods.pdf: 791527 bytes, checksum: 3cd9625f80fe5307b0d325d9a030d7a7 (MD5) https://reunir.unir.net/handle/123456789/19212 Stacked LSTM for Short-Term Wind Power Forecasting Using Multivariate Time Series Data Galphade, Manisha; Nikam, V. B.; Banerjee, Biplab; Kiwelekar, Arvind W.; Sharma, Priyanka Currently, wind power is the fast growing area in the domain of renewable energy generation. Accurate prediction of wind power output in wind farms is crucial for addressing the challenges associated the power grid. This precise forecasting enables grid operators to enhance safety and optimize grid operations by effectively managing fluctuations in power generation, ensuring a reliable and stable energy supply. In recent years, there has been a significant rise in research and investigations conducted in this field. This study aims to develop a multivariate short-term wind power forecasting (WPF) model with the objective of enhancing forecasting precision. Among the various prediction models, deep learning models such as Long Short-Term Memory (LSTM) have demonstrated outstanding performance in the field of WPF. By adding multiple layers of LSTM networks, the model can capture more complex patterns. To improve the performance, data preprocessing is carried out using two techniques such as removal of missing values and imputing missing values using Random Forest Regressor (RFR). The comparison between the proposed Stacked LSTM model and other methods including vector autoregressive (VAR), Multiple Linear Regression, Gated Recurrent Unit (GRU) and Bidirectional LSTM (BiLSTM) has been experimented on two datasets. The experimental results show that after imputing missing values using RFR, the Stacked LSTM is optimized model for better performance than above mentioned reference models. Submitted by Angela María Porras Ruiz (angela.porras@unir.net) on 2026-03-10T16:37:45Z No. of bitstreams: 1 Stacked LSTM for Short-Term Wind Power Forecasting Using Multivariate Time Series Data.pdf: 1302353 bytes, checksum: 4543961938eeff2158406c9a8ac0a471 (MD5); Made available in DSpace on 2026-03-10T16:37:45Z (GMT). No. of bitstreams: 1 Stacked LSTM for Short-Term Wind Power Forecasting Using Multivariate Time Series Data.pdf: 1302353 bytes, checksum: 4543961938eeff2158406c9a8ac0a471 (MD5) https://reunir.unir.net/handle/123456789/19211 Use of Optimised LSTM Neural Networks Pre-Trained With Synthetic Data to Estimate PV Generation Martínez Comesaña, Miguel; Martínez Torres, Javier; Javier, Pablo; López Gómez, Javier Optimising the use of the photovoltaic (PV) energy is essential to reduce fossil fuel emissions by increasing the use of solar power generation. In recent years, research has focused on physical simulations or artifical intelligence models attempting to increase the accuracy of PV generation predictions. The use of simulated data as pre-training for deep learning models has increased in different fields. The reasons are the higher efficiency in the subsequent training with real data and the possibility of not having real data available. This work presents a methodology, based on an deep learning model optimised with specific techniques and pre-trained with synthetic data, to estimate the generation of a PV system. A case study of a photovoltaic installation with 296 PV panels located in northwest Spain is presented. The results show that the model with proper pre-training trains six to seven times faster than a model without pre-training and three to four times faster than a model pre-trained with non-accurate simulated data. In terms of accuracy and considering a homogeneous training process, all models obtained average relative errors around 12%, except the model with incorrect pre-training which performs worse. Submitted by Angela María Porras Ruiz (angela.porras@unir.net) on 2026-03-10T16:33:56Z No. of bitstreams: 1 Use of Optimised LSTM Neural Networks Pre-Trained With Synthetic Data to Estimate PV Generation.pdf: 874414 bytes, checksum: e985093a6df3c4486e1d497ba70f6f66 (MD5); Made available in DSpace on 2026-03-10T16:33:56Z (GMT). No. of bitstreams: 1 Use of Optimised LSTM Neural Networks Pre-Trained With Synthetic Data to Estimate PV Generation.pdf: 874414 bytes, checksum: e985093a6df3c4486e1d497ba70f6f66 (MD5) https://reunir.unir.net/handle/123456789/19210 Trends in Addiction to Psychoactive Substances Among Homeless People in Colombia Using Artificial Intelligence Ordoñez, Hugo; Timarán Pereira, Ricardo; González Sanabria, Juan Sebastián Introduction: Currently, homelessness should not be seen as just another problem, but as a reality of inequality and the absence of social justice. In this sense, homeless people are subjected to social disengagement, lack of job opportunities or the instability of these, insecurity circumstances, these aspects being one of the causes associated with the consumption or addiction to psychoactive substances. Data: To define the proposed approach, data from the Census of Street Inhabitants - CHC- 2021 of the National Administrative Department of Statistics (DANE), which contains 19,375 records and 25 columns, were used. Methodology: This article presents an artificial intelligence approach that implements a model based on machine learning algorithms for identifying addiction trends to psychoactive substances in street dwellers in Colombia. Conclusions: Based on the results obtained, it is evident that the approach can serve as a support for decision making by municipal administrations in the definition of social public policies for the street-dwelling population in Colombia. Submitted by Angela María Porras Ruiz (angela.porras@unir.net) on 2026-03-10T16:30:02Z No. of bitstreams: 1 Trends in Addiction to Psychoactive Substances Among Homeless People in Colombia Using Artificial Intelligence.pdf: 565233 bytes, checksum: b4e896828cd0cfad07a43ea55b209154 (MD5); Made available in DSpace on 2026-03-10T16:30:02Z (GMT). No. of bitstreams: 1 Trends in Addiction to Psychoactive Substances Among Homeless People in Colombia Using Artificial Intelligence.pdf: 565233 bytes, checksum: b4e896828cd0cfad07a43ea55b209154 (MD5) https://reunir.unir.net/handle/123456789/19209 Explainable Artificial Intelligence-Based Diseases Diagnosis From Unstructured Clinical Data and Decision Making Using Blockchain Technologies M., Sumathi; Raja, S.P. In the digital era, health information is stored in digital form for easy maintenance, analysis and transfer. The proficiency of manual illness diagnosis and drug prediction in the medical field depends on the expertise availability, and experience of the specialists. In emergency and abnormal situation, the patient’s life completely depends on expert’s availability. Therefore, a different approach is needed to get around the difficulties in managing emergency cases. Artificial intelligence helps to take decisions in an accurate manner but does not provide the details of the decisions. The ability to treat emergency patients entirely depends on the particular hospitals. The clinical data includes numerical results, text prescriptions, scanned images, etc. Therefore, managing unstructured data with care is necessary for making clinical decisions. An explainable artificial intelligence-based disease diagnosis and blockchain-based decision-making system are presented in this work to address these challenges and improve patient care. A natural language processing system analyzes the unstructured data to identify different types of data and explainable AI diagnosis disease with justification and reason for the prediction. An ant colony optimization-based recommender system examines the predicted decision and identifies the specific drug for the disease. The disease decision and drug information are kept in a permissioned blockchain for confirmation. Decisions are validated by more than 50% of the experts present in the permissioned blockchain network, which consists of experts from various regions. As a result, the quickest and most accurate decisions possible are taken to handle emergency situations. Submitted by Angela María Porras Ruiz (angela.porras@unir.net) on 2026-03-10T16:26:30Z No. of bitstreams: 1 Explainable Artificial Intelligence-Based Diseases.pdf: 755211 bytes, checksum: 7b0d1912a5c5a87d6259ed890b6da20c (MD5); Made available in DSpace on 2026-03-10T16:26:30Z (GMT). No. of bitstreams: 1 Explainable Artificial Intelligence-Based Diseases.pdf: 755211 bytes, checksum: 7b0d1912a5c5a87d6259ed890b6da20c (MD5) https://reunir.unir.net/handle/123456789/19208 Testing Deep Learning Recommender Systems Models on Synthetic GAN-Generated Datasets Bobadilla, Jesús; Gutiérrez, Abraham The published method Generative Adversarial Networks for Recommender Systems (GANRS) allows generating data sets for collaborative filtering recommendation systems. The GANRS source code is available along with a representative set of generated datasets. We have tested the GANRS method by creating multiple synthetic datasets from three different real datasets taken as a source. Experiments include variations in the number of users in the synthetic datasets, as well as a different number of samples. We have also selected six state-of-the-art collaborative filtering deep learning models to test both their comparative performance and the GANRS method. The results show a consistent behavior of the generated datasets compared to the source ones; particularly, in the obtained values and trends of the precision and recall quality measures. The tested deep learning models have also performed as expected on all synthetic datasets, making it possible to compare the results with those obtained from the real source data. Future work is proposed, including different cold start scenarios, unbalanced data, and demographic fairness. Submitted by Angela María Porras Ruiz (angela.porras@unir.net) on 2026-03-10T16:22:47Z No. of bitstreams: 1 Testing Deep Learning Recommender Systems Models on Synthetic GAN-Generated Datasets.pdf: 1472576 bytes, checksum: 2860c05b019e926dcea84f05f95619de (MD5); Made available in DSpace on 2026-03-10T16:22:47Z (GMT). No. of bitstreams: 1 Testing Deep Learning Recommender Systems Models on Synthetic GAN-Generated Datasets.pdf: 1472576 bytes, checksum: 2860c05b019e926dcea84f05f95619de (MD5) https://reunir.unir.net/handle/123456789/19207 Large Language Models for in Situ Knowledge Documentation and Access With Augmented Reality Izquierdo Domenech, Juan; Linares Pellicer, Jordi; Ferri Molla, Isabel Augmented reality (AR) has become a powerful tool for assisting operators in complex environments, such as shop floors, laboratories, and industrial settings. By displaying synthetic visual elements anchored in real environments and providing information for specific tasks, AR helps to improve efficiency and accuracy. However, a common bottleneck in these environments is introducing all necessary information, which often requires predefined structured formats and needs more ability for multimodal and Natural Language (NL) interaction. This work proposes a new method for dynamically documenting complex environments using AR in a multimodal, non-structured, and interactive manner. Our method employs Large Language Models (LLMs) to allow experts to describe elements from the real environment in NL and select corresponding AR elements in a dynamic and iterative process. This enables a more natural and flexible way of introducing information, allowing experts to describe the environment in their own words rather than being constrained by a predetermined structure. Any operator can then ask about any aspect of the environment in NL to receive a response and visual guidance from the AR system, thus allowing for a more natural and flexible way of introducing and retrieving information. These capabilities ultimately improve the effectiveness and efficiency of tasks in complex environments. Submitted by Angela María Porras Ruiz (angela.porras@unir.net) on 2026-03-10T16:18:55Z No. of bitstreams: 1 Large Language Models for in Situ Knowledge Documentation and Access With Augmented Reality.pdf: 1472569 bytes, checksum: 751baa5d6dfe340da811523a6afd0c1e (MD5); Made available in DSpace on 2026-03-10T16:18:55Z (GMT). No. of bitstreams: 1 Large Language Models for in Situ Knowledge Documentation and Access With Augmented Reality.pdf: 1472569 bytes, checksum: 751baa5d6dfe340da811523a6afd0c1e (MD5) https://reunir.unir.net/handle/123456789/19206 Distinguishing Human From Machine: A Review of Advances and Challenges in AI-Generated Text Detection Fariello, Serena The rise of Large Language Models (LLMs) has dramatically altered the generation and spreading of textual content. This advancement offers benefits in various domains, including medicine, education, law, coding, and journalism, but also has negative implications, mainly related to ethical concerns. Preventing measures to mitigate negative implications pass through solutions that distinguish machine-generated text from humanwritten text. This study aims to provide a comprehensive review of existing literature for detecting LLMgenerated texts. Emerging techniques are categorized into five categories: watermarking, feature-based, neural-based, hybrid, and human-aided methods. For each introduced category, strengths and limitations are discussed, providing insights into their effectiveness and potential for future improvements. Moreover, available datasets and tools are introduced. Results demonstrate that, despite the good delimited performance, the multitude of languages to recognize, hybrid texts, the continuous improvement of algorithms for text generation and the lack of regulation require additional efforts for efficient detection. Submitted by Angela María Porras Ruiz (angela.porras@unir.net) on 2026-03-10T16:12:44Z No. of bitstreams: 1 Distinguishing Human From Machine A Review of Advances and Challenges in AI-Generated Text Detection.pdf: 465865 bytes, checksum: 9b07ebd124a70095cc94411c6a5a1a6d (MD5); Made available in DSpace on 2026-03-10T16:12:44Z (GMT). No. of bitstreams: 1 Distinguishing Human From Machine A Review of Advances and Challenges in AI-Generated Text Detection.pdf: 465865 bytes, checksum: 9b07ebd124a70095cc94411c6a5a1a6d (MD5) https://reunir.unir.net/handle/123456789/19205 Editor’s Note Morente Molinera, Juan Antonio The International Journal of Interactive Multimedia and Artificial Intelligence (IJIMAI) publishes articles discussing the latest current topics in the research literature. The emergence of ChatGPT and other similar models based on deep learning are dramatically changing the way people understand and use artificial intelligence. Despite the significant advances made in these types of techniques, which have been enormous in recent years, new learning methods are still needed. Specifically, we require methods that allow us to handle data correctly in specific environments, as well as provide learning methods with the necessary explainability that allows us to understand how they are reasoning. The latter is essential for creating ethical learning methods that do not make unfair decisions based on biased information. It is also important to identify data that have, in some way, reflected the reprehensible attitudes and reasoning that we as fallible human beings sometimes have. In short, artificial intelligence should reflect, if possible, the best of us rather than the worst. With this goal in mind, it is common to see in this issue of the journal an abundance of articles proposing new learning methods, many of which are based on Deep Learning and Data Mining. There are also articles on large language models, which are extremely important in the current artificial intelligence landscape. Of course, there are also articles on optimization methods and quantum computers, which are also of great importance in the field of artificial intelligence. Although generative artificial intelligence models are perhaps the ones that have people most intrigued, this is not the only current application of artificial intelligence. We are seeing how renewable energies, in particular those that come from the sun and wind, are playing an increasingly important role in global energy generation. As seen in recent events, such as the general blackout in Spain, the electricity system needs new methods that allow adequate regulation to prevent all kinds of possible failures. In this issue, two articles present new applications of artificial intelligence methods to renewable energy generation systems. Also noteworthy within this issue is the application of artificial intelligence in the field of teaching, where the aim is to provide a better learning experience for students and teachers. Submitted by Angela María Porras Ruiz (angela.porras@unir.net) on 2026-03-10T16:09:26Z No. of bitstreams: 1 Editor’s Note.pdf: 72460 bytes, checksum: d61494befc2e8cee6c6f9678d703da8b (MD5); Made available in DSpace on 2026-03-10T16:09:26Z (GMT). No. of bitstreams: 1 Editor’s Note.pdf: 72460 bytes, checksum: d61494befc2e8cee6c6f9678d703da8b (MD5) https://reunir.unir.net/handle/123456789/19202 Platform for Improving the User Experience in the Creation of Educational Multiplayer Video Games Sánchez Canella, Fernando; Pascual Espada, Jordán; Cid Rico, Irene Students’ motivation is one of the factors that directly affect academic performance. In recent years, teachers are looking for ways to motivate students during their training period. For example, making use of slides, videos, films, comics or games to increase students' motivation to improve their learning experience. Some research works have revealed that multiplayer games which include cooperation and competition, among other factors, are an extraordinary tool for enhancing students’ motivation. Current alternatives make it very complex for teachers to create multiplayer games for their students. The definition of the game requires many configurations and even technical knowledge. This research proposes a new platform that allows teachers to create multiplayer video games in a simple and fast way, improving the game creation process over current alternatives. The resulting games are also designed for to improve the student experience, and make it fun. These games do not only include trivia questions, but also use functional mechanisms from video games. The design of the generated games allows students to master the games in a short period of time during their classes. Submitted by Angela María Porras Ruiz (angela.porras@unir.net) on 2026-03-10T15:47:39Z No. of bitstreams: 1 Platform for Improving the User Experience in the Creation of Educational Multiplayer Video Games.pdf: 370917 bytes, checksum: 8ad91deb52bfcab9d34dacd6bc8afe31 (MD5); Made available in DSpace on 2026-03-10T15:47:39Z (GMT). No. of bitstreams: 1 Platform for Improving the User Experience in the Creation of Educational Multiplayer Video Games.pdf: 370917 bytes, checksum: 8ad91deb52bfcab9d34dacd6bc8afe31 (MD5) https://reunir.unir.net/handle/123456789/19201 Selecting the Appropriate User Experience Questionnaire and Guidance for Interpretation: the UEQ Family Kollmorgen, Jessica; Hinderks, Andreas; Thomaschewski, Jörg Measuring the user experience (UX) of products, systems and services is individual depending on the research question. On the one hand, the user’s goals and environment play a role in the subjective evaluation. On the other hand, different UX factors are relevant depending on the product. In this case, it is practical to have a questionnaire family as an aid, whose questionnaires are geared towards these different use cases. The User Experience Questionnaire (UEQ) family allows researchers and practitioners to choose the right tool for efficient UX measurement from three questionnaire versions. This article summarizes the UEQ, its short version (UEQ-S) and a modular framework (UEQ+) with overall 27 UX factors and purposes in over 30 different languages. In addition, specific instructions and assistance are provided for the statistical evaluation and interpretation of the questionnaire results. With the help of a key performance indicator (KPI), benchmarks and an importance-performance analysis (IPA), the realization of UX measurements is made easier for researchers and practitioners. To make it even more convenient to choose the right questionnaire from the UEQ family, influencing factors on the UX measurement and recommendations for action are given. Submitted by Angela María Porras Ruiz (angela.porras@unir.net) on 2026-03-10T15:44:06Z No. of bitstreams: 1 Selecting the Appropriate User Experience Questionnaire and Guidance for Interpretation.pdf: 434063 bytes, checksum: 002e410b3900c0fb3c7080e0981b38bd (MD5); Made available in DSpace on 2026-03-10T15:44:06Z (GMT). No. of bitstreams: 1 Selecting the Appropriate User Experience Questionnaire and Guidance for Interpretation.pdf: 434063 bytes, checksum: 002e410b3900c0fb3c7080e0981b38bd (MD5) https://reunir.unir.net/handle/123456789/19200 Combating Misinformation and Polarization in the Corporate Sphere: Integrating Social, Technological and AI Strategies Tejero, Alberto; Pisoni, Galena; Lashkari, Ziba Habibi; Rios Aguilar, Sergio In an era where misinformation and polarization present significant challenges, this research examines the root causes within social networks and assesses how corporations can use AI technologies for prompt detection. This research uses a dual approach: a "telephone game" with 225 participants from a Spanish university to study the spread of misinformation, and interviews with 15 experts from three French tech companies to investigate technological solutions. The findings indicate that almost one-third of participants inadvertently contribute to polarization, and around one-quarter propagated misinformation. The study also identifies the existing tools enhanced by AI and Machine Learning that effectively detect misinformation and polarization in corporate settings. This investigation provides crucial insights for practitioners to strengthen their strategies against misinformation and technical challenges and opportunities. Submitted by Angela María Porras Ruiz (angela.porras@unir.net) on 2026-03-10T15:39:48Z No. of bitstreams: 1 Combating Misinformation and Polarization in the Corporate Sphere.pdf: 309075 bytes, checksum: db6ffe3ce4ba578fd9d7392e4facc16b (MD5); Made available in DSpace on 2026-03-10T15:39:48Z (GMT). No. of bitstreams: 1 Combating Misinformation and Polarization in the Corporate Sphere.pdf: 309075 bytes, checksum: db6ffe3ce4ba578fd9d7392e4facc16b (MD5) https://reunir.unir.net/handle/123456789/19198 Automatic Surveillance of People and Objects on Railway Tracks Martínez Núñez, Domingo; López Hernández, Fernando Carlos; Rainer Granados, J. Javier This paper describes the development and evaluation of a surveillance system for the detection of people and objects on railroad tracks in real time. Firstly, the paper evaluates several background subtraction techniques including CNNs and the object detection library called YOLO. Then we describe a novel strategy to mitigate the occlusion caused by the perspective of the camera and the integration of an alarms and pre-alarms policy. To evaluate its performance, we have implemented and automated the control and notification aspects of the surveillance system using computer vision techniques. This setup, running on a standard PC, achieves an average frame rate of 15 FPS and a latency of 0.54 seconds per frame, meeting real-time expectations in terms of both false alarms and precision in operational mode. The results from experiments conducted with a publicly available recorded video dataset from Metro de Madrid facilities demonstrate significant improvements over current state-of the-art solutions. These improvements include better accident anticipation and enhanced information provided to the operator using a standard low-cost camera. Consequently, we conclude that the approach described in this paper is both effective and a more practical, cost-efficient alternative to the other solutions reviewed. Submitted by Angela María Porras Ruiz (angela.porras@unir.net) on 2026-03-10T15:10:54Z No. of bitstreams: 1 Automatic Surveillance of People and Objects on Railway Tracks.pdf: 471901 bytes, checksum: 95951b5a4386c4fbc12f3570d95c43f2 (MD5); Made available in DSpace on 2026-03-10T15:10:54Z (GMT). No. of bitstreams: 1 Automatic Surveillance of People and Objects on Railway Tracks.pdf: 471901 bytes, checksum: 95951b5a4386c4fbc12f3570d95c43f2 (MD5) https://reunir.unir.net/handle/123456789/19197 Multiscale Attentional Squeeze-And-Excitation Network for Person Re-Identification Guo, Tiancun; Zhou, Qiang; Gao, Mingliang; Jeon, Gwanggil; Camacho, David In recent years, with the advancement of deep learning, person re-identification (Re-ID) has become increasingly significant. The existing person Re-ID methods primarily focus on optimizing network architecture to enhance Re-ID task performance. However, these methods often overlook the importance of valuable features in distinguishing Re-ID tasks, leading to reduced model efficacy in complex scenarios. As a solution, we utilize the attention mechanism to develop the lightweight multiscale Attentional Squeeze-and-Excitation Network (MASENet) that can distinguish between significant and non-significant features. Specifically, we utilize the SEAttention (SE) module to amplify important feature channels and suppress redundant ones. Additionally, the Spatial Group Enhance (SGE) module is introduced to enable networks to enhance semantic learning expression and suppress potential noise autonomously. We conduct comprehensive experiments on Market1501, MSMT17, and VeRi-776 datasets and cross-domain experiments on MSMT17 Ñ Market1501 to validate the model performance. Experimental results prove that the proposed MASENet achieves competitive performance across all experiments. Submitted by Angela María Porras Ruiz (angela.porras@unir.net) on 2026-03-10T15:01:51Z No. of bitstreams: 1 Multiscale Attentional Squeeze-And-Excitation Network for Person Re-Identification.pdf: 736449 bytes, checksum: 34ea53121002bf315ec3e13ad6571854 (MD5); Made available in DSpace on 2026-03-10T15:01:51Z (GMT). No. of bitstreams: 1 Multiscale Attentional Squeeze-And-Excitation Network for Person Re-Identification.pdf: 736449 bytes, checksum: 34ea53121002bf315ec3e13ad6571854 (MD5) https://reunir.unir.net/handle/123456789/19196 Prediction of COVID-19 Using a Clinical Dataset With Machine Learning Approaches Suruliandi, A.; Rayan, R. Ame; Raja, S. P. COVID-19 is an infectious disease that spreads quickly from person to another. The pandemic, which spread worldwide over time, presents huge risks in terms of blood clotting, breathing problems and heart attacks, sometimes with fatal consequences if not detected early. The PCR test, CT scans, X-rays, and blood tests are methods commonly employed to detect the disease, though the PCR test is, without question, considered the gold standard. The American Center for Disease Control and Prevention (CDC) reports that the PCR has an 80% accuracy rate. An alternative to the PCR is clinical data, which is less expensive, easy to collect, and offers better accuracy. Machine learning, with its rich feature selection and classification methods, helps detect COVID-19 at the earliest stages, using clinical test results. This research proposes a clinical dataset and offers a comparative analysis of feature selection and classification algorithms for detecting COVID-19. Filter-based feature selection methods such as the ANOVA-F, chi-square, mutual information and Pearson correlation, along with wrapperbased methods such as Recursive Feature Elimination (RFE) and Sequential Forward Selection (SFS) were used to choose a subset of features from the feature set. The selected features were thereafter applied to the Support Vector Machine (SVM), Naïve Bayes, K-NN (K-Nearest Neighbor) and Logistic Regression(LR) classification algorithms to detect Coronavirus Disease. The experimental results of the comparative study show that the clinical dataset provides better accuracy at 94.8%, with mutual information and the SVM classifier. Submitted by Angela María Porras Ruiz (angela.porras@unir.net) on 2026-03-10T14:57:49Z No. of bitstreams: 1 Prediction of COVID-19 Using a Clinical Dataset With Machine Learning Approaches.pdf: 772806 bytes, checksum: 1a4176226460cd5514e73a1c0da4a6e1 (MD5); Made available in DSpace on 2026-03-10T14:57:49Z (GMT). No. of bitstreams: 1 Prediction of COVID-19 Using a Clinical Dataset With Machine Learning Approaches.pdf: 772806 bytes, checksum: 1a4176226460cd5514e73a1c0da4a6e1 (MD5) https://reunir.unir.net/handle/123456789/19195 TKU-PSO: An Efficient Particle Swarm Optimization Model for Top-K High-Utility Itemset Mining Carstensen, Simen; Lin, Jerry Chun Wei Top-k high-utility itemset mining (top- HUIM) is a data mining procedure used to identify the most valuable patterns within transactional data. Although many algorithms are proposed for this purpose, they require substantial execution times when the search space is vast. For this reason, several meta-heuristic models have been applied in similar utility mining problems, particularly evolutionary computation (EC). These algorithms are beneficial as they can find optimal solutions without exploring the search space exhaustively. However, there are currently no evolutionary heuristics available for top-k HUIM. This paper addresses this issue by proposing an EC-based particle swarm optimization model for top-k HUIM, which we call TKU-PSO. In addition, we have developed several strategies to relieve the computational complexity throughout the algorithm. First, redundant and unnecessary candidate evaluations are avoided by utilizing explored solutions and estimating itemset utilities. Second, unpromising items are pruned during execution based on a thresholdraising concept we call minimum solution fitness. Finally, the traditional population initialization approach is revised to improve the model’s ability to find optimal solutions in huge search spaces. Our results show that TKU-PSO is faster than state-of-the-art competitors in all datasets tested. Most notably, existing algorithms could not complete certain experiments due to excessive runtimes, whereas our model discovered the correct solutions within seconds. Moreover, TKU-PSO achieved an overall accuracy of 99.8% compared to 16.5% with the current heuristic approach, while memory usage was the smallest in 2/3 of all tests. Submitted by Angela María Porras Ruiz (angela.porras@unir.net) on 2026-03-10T14:48:57Z No. of bitstreams: 1 TKU-PSO An Efficient Particle Swarm Optimization Model for Top-K High-Utility Itemset Mining.pdf: 797963 bytes, checksum: 65eb3b395f4ae5aa6076d01176e325c4 (MD5); Made available in DSpace on 2026-03-10T14:48:57Z (GMT). No. of bitstreams: 1 TKU-PSO An Efficient Particle Swarm Optimization Model for Top-K High-Utility Itemset Mining.pdf: 797963 bytes, checksum: 65eb3b395f4ae5aa6076d01176e325c4 (MD5) https://reunir.unir.net/handle/123456789/19194 Optimal Target-Oriented Knowledge Transportation For Aspect-Based Multimodal Sentiment Analysis Zhang, Linhao; Jin, Li; Xu, Guangluan; Li, Xiaoyu; Sun, Xian; Zhang, Zequn; Zhang, Yanan Aspect-based multimodal sentiment analysis under social media scenario aims to identify the sentiment polarities of each aspect term, which are mentioned in a piece of multimodal user-generated content. Previous approaches for this interdisciplinary multimodal task mainly rely on coarse-grained fusion mechanisms from the data-level or decision-level, which have the following three shortcomings:(1) ignoring the category knowledge of the sentiment target mentioned in the text) in visual information. (2) unable to assess the importance of maintaining target interaction during the unimodal encoding process, which results in indiscriminative representations considering various aspect terms. (3) suffering from the semantic gap between multiple modalities. To tackle the above challenging issues, we propose an optimal target-oriented knowledge transportation network (OtarNet) for this task. Firstly, the visual category knowledge is explicitly transported through input space translation and reformulation. Secondly, with the reformulated knowledge containing the target and category information, the target sensitivity is well maintained in the unimodal representations through a multistage target-oriented interaction mechanism. Finally, to eliminate the distributional modality gap by integrating complementary knowledge, the target-sensitive features of multiple modalities are implicitly transported based on the optimal transport interaction module. Our model achieves state-of-theart performance on three benchmark datasets: Twitter-15, Twitter-17 and Yelp, together with the extensive ablation study demonstrating the superiority and effectiveness of OtarNet. Submitted by Angela María Porras Ruiz (angela.porras@unir.net) on 2026-03-10T14:44:20Z No. of bitstreams: 1 Optimal Target-Oriented Knowledge Transportation For Aspect-Based Multimodal Sentiment Analysis.pdf: 1152006 bytes, checksum: 28abb3788caafe9c124f2fc62b97bf5d (MD5); Made available in DSpace on 2026-03-10T14:44:20Z (GMT). No. of bitstreams: 1 Optimal Target-Oriented Knowledge Transportation For Aspect-Based Multimodal Sentiment Analysis.pdf: 1152006 bytes, checksum: 28abb3788caafe9c124f2fc62b97bf5d (MD5) https://reunir.unir.net/handle/123456789/19192 A Robust Framework for Speech Emotion Recognition Using Attention Based Convolutional Peephole LSTM Paramasivam, Ramya; Lavanya, K.; Divakarachari, Parameshachari Bidare; Camacho, David Speech Emotion Recognition (SER) plays an important role in emotional computing which is widely utilized in various applications related to medical, entertainment and so on. The emotional understanding improvises the user machine interaction with a better responsive nature. The issues faced during SER are existence of relevant features and increased complexity while analyzing of huge datasets. Therefore, this research introduces a wellorganized framework by introducing Improved Jellyfish Optimization Algorithm (IJOA) for feature selection, and classification is performed using Convolutional Peephole Long Short-Term Memory (CP-LSTM) with attention mechanism. The raw data acquisition takes place using five datasets namely, EMO-DB, IEMOCAP, RAVDESS, Surrey Audio-Visual Expressed Emotion (SAVEE) and Crowd-sourced Emotional Multimodal Actors Dataset (CREMA-D). The undesired partitions are removed from the audio signal during pre-processing and fed into phase of feature extraction using IJOA. Finally, CP LSTM with attention mechanisms is used for emotion classification. As the final stage, classification takes place using CP-LSTM with attention mechanisms. Experimental outcome clearly shows that the proposed CP-LSTM with attention mechanism is more efficient than existing DNN-DHO, DH-AS, D-CNN, CEOAS methods in terms of accuracy. The classification accuracy of the proposed CP-LSTM with attention mechanism for EMO-DB, IEMOCAP, RAVDESS and SAVEE datasets are 99.59%, 99.88%, 99.54% and 98.89%, which is comparably higher than other existing techniques. Submitted by Angela María Porras Ruiz (angela.porras@unir.net) on 2026-03-10T13:02:22Z No. of bitstreams: 1 A Robust Framework for Speech Emotion Recognition Using Attention Based Convolutional Peephole LSTM.pdf: 702337 bytes, checksum: 0a1e512c40ee7c108ce218b4af7828a0 (MD5); Made available in DSpace on 2026-03-10T13:02:22Z (GMT). No. of bitstreams: 1 A Robust Framework for Speech Emotion Recognition Using Attention Based Convolutional Peephole LSTM.pdf: 702337 bytes, checksum: 0a1e512c40ee7c108ce218b4af7828a0 (MD5) https://reunir.unir.net/handle/123456789/19156 An Adaptive Salp-Stochastic-Gradient-Descent-Based Convolutional LSTM With MapReduce Framework for the Prediction of Rainfall Manoj, S. Oswalt; Kumar, Abhishek; Dubey, Ashutosh Kumar; Ananth, J. P. Rainfall prediction is considered to be an esteemed research area that impacts the day-to-day life of Indians. The predominant income source of most of the Indian population is agriculture. It helps the farmers to make the appropriate decisions pertaining to cultivation and irrigation. The primary objective of this investigation is to develop a technique for rainfall prediction utilising the MapReduce framework and the convolutional long short-term memory (ConvLSTM) method to circumvent the limitations of higher computational requirements and the inability to process a large number of data points. In this work, an adaptive salp-stochastic-gradientdescent-based ConvLSTM (adaptive S-SGD-based ConvLSTM) system has been developed to predict rainfall accurately to process the long time series data and to eliminate the vanishing problems. To optimize the hyperparameter of the convLSTM model, the S-SGD methodology proposed combine the SGD and the salp swarm algorithm (SSA). The adaptive S-SGD based ConvLSTM has been developed by integrating the adaptive concept in S-SGD. It tunes the weights of ConvLSTM optimally to achieve better prediction accuracy. Assessment measures, such as the percentage root mean square difference (PRD) and mean square error (MSE), were employed to compare the suggested method with previous approaches. The developed system demonstrates high prediction accuracy, achieving minimal values for MSE (0.0042) and PRD (0.8450). Submitted by Angela María Porras Ruiz (angela.porras@unir.net) on 2026-03-09T16:56:05Z No. of bitstreams: 1 An Adaptive Salp-Stochastic-Gradient-Descent-Based Convolutional LSTM With MapReduce Framework for the Prediction of Rainfall.pdf: 840750 bytes, checksum: ab83fa3761d1e8fbb514b1de528ecfe6 (MD5); Made available in DSpace on 2026-03-09T16:56:05Z (GMT). No. of bitstreams: 1 An Adaptive Salp-Stochastic-Gradient-Descent-Based Convolutional LSTM With MapReduce Framework for the Prediction of Rainfall.pdf: 840750 bytes, checksum: ab83fa3761d1e8fbb514b1de528ecfe6 (MD5) https://reunir.unir.net/handle/123456789/19155 Measuring the Difference Between Pictures From Controlled and Uncontrolled Sources to Promote a Destination. A Deep Learning Approach Diaz Pacheco, AngeL; Álvarez Carmona, Miguel A.; Rodríguez González, Ansel Y.; Carlos, Hugo; Aranda, Ramón Promoting a destination is a major task for Destination Marketing Organizations (DMOs). Although DMOs control, to some extent, the information presented to travelers (controlled sources), there are other different sources of information (uncontrolled sources) that could project an unfavorable image of the destination. Measuring differences between information sources would help design strategies to mitigate negative factors. In this way, we propose a deep learning-based approach to automatically measure the changes between images from controlled and uncontrolled information sources. Our approach exempts experts from the time-consuming task of assessing enormous quantities of pictures to track changes. To our best knowledge, this work is the first work that focuses on this issue using technological paradigms. Notwithstanding this, our approach paves novel pathways to acquire strategic insights that can be harnessed for the augmentation of destination development, the refinement of recommendation systems, the analysis of online travel reviews, and myriad other pertinent domains. Submitted by Angela María Porras Ruiz (angela.porras@unir.net) on 2026-03-09T16:52:08Z No. of bitstreams: 1 Measuring the Difference Between Pictures From Controlled and Uncontrolled Sources to Promote a Destination. A Deep Learning Approach.pdf: 957662 bytes, checksum: 14760a4d6ccbf933247a0c428df68e6a (MD5); Made available in DSpace on 2026-03-09T16:52:08Z (GMT). No. of bitstreams: 1 Measuring the Difference Between Pictures From Controlled and Uncontrolled Sources to Promote a Destination. A Deep Learning Approach.pdf: 957662 bytes, checksum: 14760a4d6ccbf933247a0c428df68e6a (MD5) https://reunir.unir.net/handle/123456789/19154 Performance and Communication Cost of Deep Neural Networks in Federated Learning Environments: An Empirical Study Alotaibi, Basmah K.; Khan, Fakhri Alam; Qawqzeh, Yousef; Jeon, Gwanggil; Camacho, David Federated learning, a distributive cooperative learning approach, allows clients to train the model locally using their data and share the trained model with a central server. When developing a federated learning environment, a deep/machine learning model needs to be chosen. The choice of the learning model can impact the model performance and the communication cost since federated learning requires the model exchange between clients and a central server in several rounds. In this work, we provide an empirical study to investigate the impact of using three different neural networks (CNN, VGG, and ResNet) models in image classification tasks using two different datasets (Cifar-10 and Cifar-100) in a federated learning environment. We investigate the impact of using these models on the global model performance and communication cost under different data distribution that are IID data and non-IID data distribution. The obtained results indicate that using CNN and ResNet models provide a faster convergence than VGG model. Additionally, these models require less communication costs. In contrast, the VGG model necessitates the sharing of numerous bits over several rounds to achieve higher accuracy under the IID data settings. However, its accuracy level is lower under non-IID data distributions than the other models. Furthermore, using a light model like CNN provides comparable results to the deeper neural network models with less communication cost, even though it may require more communication rounds to achieve the target accuracy in both datasets. CNN model requires fewer bits to be shared during communication than other models. Submitted by Angela María Porras Ruiz (angela.porras@unir.net) on 2026-03-09T16:48:22Z No. of bitstreams: 1 Performance and Communication Cost of Deep Neural Networks in Federated Learning Environments An Empirical Study.pdf: 794820 bytes, checksum: f3c58e1d4566710f4a58038bd98a794e (MD5); Made available in DSpace on 2026-03-09T16:48:22Z (GMT). No. of bitstreams: 1 Performance and Communication Cost of Deep Neural Networks in Federated Learning Environments An Empirical Study.pdf: 794820 bytes, checksum: f3c58e1d4566710f4a58038bd98a794e (MD5) https://reunir.unir.net/handle/123456789/19153 Editor’s Note García Martínez Eyre i Canals, Yamila Artificial Intelligence (AI) is a scientific discipline that aims to drive disruptive scenarios for science-based technical developments that solve complex problems. The IJIMAI journal’s scope is precisely to demonstrate how the combination of two factors — technical foundations and sought-after applications — must guide future AI developments to find solutions to complex real-world problems. This IJIMAI publication opens with an article that considers the current framework for AI fundamentals: how can we improve AI technology to find solutions to real-unsolved problems? The initial answer seems to be related with a desired self-consistent procedure: let machines learn from our experience. In the article by Alotaibi et al., the analysis of neural networks in terms of the parameters used, how they work, and how do they respond to the problem itself led the authors to a rationale for decision-making regarding the performance of different neural models. The immediate question that arises is whether there are any universal and fundamental criteria that can be used to define the models that guide AI methods. Apparently, there are not such universal methods, and we are faced with a challenging open problem. Subsequent manuscripts will provide readers with more in-depth insights into this issue. Submitted by Angela María Porras Ruiz (angela.porras@unir.net) on 2026-03-09T16:43:57Z No. of bitstreams: 1 Editor’s Note.pdf: 72460 bytes, checksum: d61494befc2e8cee6c6f9678d703da8b (MD5); Made available in DSpace on 2026-03-09T16:43:57Z (GMT). No. of bitstreams: 1 Editor’s Note.pdf: 72460 bytes, checksum: d61494befc2e8cee6c6f9678d703da8b (MD5) https://reunir.unir.net/handle/123456789/19132 Security Model for the Internet of Things, Through Blockchain Díaz Gutiérrez, Yesid; Cueva-Lovelle, Juan Manuel; Candia Herrera, Diana Carolina Due to the proliferation of computer crimes related to information vulnerability handled by people and entities and evidenced in attacks of financial, commercial, personal and even family nature; a need has been identified to implement, security strategies and protocols in each and every one of these areas, which make possible the effective protection of the integrity and privacy of data. Regarding this, there are protection schemes such as cryptography and reliable time stamping which undoubtedly have managed to partially solve this problem by attacking structural and crucial points. However, the evolution in the technology field has been currently represented in the fourth industrial revolution and its context towards 4.0 technologies and smart industries; various technologies have been positioned in the emerging and disruptive categories, among which the Internet of Things (IoT) stands out. This technology has become the target of multiple computer attacks, due to the processes of Extraction, Transformation, Loading and Transmission of large volumes of data. Alongside its widespread connection to the Internet, it’s become a strategic target for such attacks. A possible alternative solution to this situation is blockchain, which allows information to be public and stored in different blocks, which makes it easier to guarantee the integrity of information based on the following aspects: • Identification of the attacked and / or compromised information, which can be marked as invalid information. • Public report of the attack. • Information backup in another block to facilitate its recovery. In this regard, it is important to highlight that these functional and technological characteristics offered by the blockchain, facilitate the management of information and its integrity. However, it is necessary and essential to previously guarantee the structure of the information generated; as some processes of Business Intelligence (BI), such as the Extraction, Transformation and Load scheme (ELT), would be of great relevance and support during the development of this procedure. Submitted by Angela María Porras Ruiz (angela.porras@unir.net) on 2026-03-06T12:25:21Z No. of bitstreams: 1 Security Model for the Internet of Things, Through Blockchain.pdf: 450339 bytes, checksum: 6744590dde281fb739e033db5658a521 (MD5); Made available in DSpace on 2026-03-06T12:25:21Z (GMT). No. of bitstreams: 1 Security Model for the Internet of Things, Through Blockchain.pdf: 450339 bytes, checksum: 6744590dde281fb739e033db5658a521 (MD5) https://reunir.unir.net/handle/123456789/19131 An Adaptive Framework for Resource Allocation Management in 5G Vehicular Networks Vijayan, Rajilal Manathala; Granelli, Fabrizio; Umamakeswari, A. Vehicle-to-everything (V2X) communication is crucial in vehicular networks, for enhancing traffic safety by ensuring dependable and low latency services. However, interference has a significant impact on V2X communication when channel states are changed in a high mobility environment. Integration of next generation cellular networks such as 5G in V2X communication can solve this issue. Also, successful resource allocation among users achieves a better interference control in high mobility scenarios. This work proposes a novel resource allocation strategy for 5G cellular V2X communication based on clustering technique and Deep Reinforcement Learning (DRL) with the aim of maximizing systems energy efficiency and MVNO’s profit. DRL is used to distribute communication resources for the best interference control in high mobility scenarios. To reduce signalling overhead in DRL deployments, the proposed method adopted RRH grouping and vehicle clustering technique. The overall architecture is implemented in two phases. The first phase addresses the RRH grouping and vehicle clustering technique with the objective of maximising the energy efficiency of the system and the second phase addresses the technique of employing DRL in conjunction with bidding to optimise MVNO’s profit. Second phase addresses the resource allocation which is implemented in two level stage. First level addresses the bidding of resources to BS using bidding and DRL techniques and the second level addresses the resource allocation to users using Dueling DQN technique. Through simulations, the proposed algorithm's performance is compared with the existing algorithms and the results depicts the improved performance of the proposed system. Submitted by Angela María Porras Ruiz (angela.porras@unir.net) on 2026-03-06T12:20:20Z No. of bitstreams: 1 An Adaptive Framework for Resource Allocation Management in 5G Vehicular Networks.pdf: 1076396 bytes, checksum: 3309ad5626d22e3cdb0b884466631f03 (MD5); Made available in DSpace on 2026-03-06T12:20:20Z (GMT). No. of bitstreams: 1 An Adaptive Framework for Resource Allocation Management in 5G Vehicular Networks.pdf: 1076396 bytes, checksum: 3309ad5626d22e3cdb0b884466631f03 (MD5) https://reunir.unir.net/handle/123456789/19130 Identification of Monkeypox Disease Based on MpoxNet and Swin Transformer Models Using Mobile Application Sadesh, S.; Thangaraj, Rajasekaran; Pandiyan, P.; Priya, R. Devi; Naveen, Palanichamy Humankind is still reeling from the devastating impact of the Covid-19 pandemic, yet another looming threatis the potential global spread of the monkeypox virus. While monkeypox may not pose the same level of lethality or contagion as COVID-19, its significant spread across countries is cause for concern. Already, outbreaks have been reported in 75 nations worldwide. Despite sharing clinical characteristics with smallpox, including lesions and rashes, monkeypox symptoms are frequently mistaken for those of other poxviruses such as chickenpox and cowpox. Consequently, accurate early diagnosis of monkeypox by healthcare professionals remains challenging. Automated monkeypox identification using Deep Learning (DL) techniques presents a promising avenue for addressing this challenge. In this study, a modified deep convolutional neural network (DCNN) model named MpoxNet is proposed for the identification of monkeypox disease. The performance of MpoxNet is evaluated against established DCNN models, including ResNet50, VGG16, VGG19, DenseNet121, DenseNet169, Xception, InceptionResNetV2, and MobileNetV2. This study addresses the pressing challenge of monkeypox identification by proposing MpoxNet. With the aim of enhancing early detection and containment efforts, MpoxNet's performance is evaluated against established DCNN models across two distinct datasets: MSLD and MSID Dataset. Results reveal MpoxNet's superior test accuracy of 94.82% on the MSLD Dataset, surpassing other models. However, evaluation on the MSID Dataset highlights variations in performance, emphasizing the influence of dataset characteristics. Additionally, the introduction of the Swin Transformer model demonstrates exceptional performance on the MSLD and the MSID Dataset and, achieving an accuracy of 98%. These findings underscore the importance of considering diverse datasets and leveraging advanced techniques for robust monkeypox detection systems. Integration of MpoxNet with a mobile application offers a promising solution for rapid and precise monkeypox disease detection, providing valuable insights for future research and real-world deployment strategies to effectively combat the global spread of monkeypox. Submitted by Angela María Porras Ruiz (angela.porras@unir.net) on 2026-03-06T12:15:38Z No. of bitstreams: 1 Identification of Monkeypox Disease Based on MpoxNet and Swin Transformer Models Using Mobile Application.pdf: 783729 bytes, checksum: c12c03c73eb88128014adcf20f2337ba (MD5); Made available in DSpace on 2026-03-06T12:15:38Z (GMT). No. of bitstreams: 1 Identification of Monkeypox Disease Based on MpoxNet and Swin Transformer Models Using Mobile Application.pdf: 783729 bytes, checksum: c12c03c73eb88128014adcf20f2337ba (MD5) https://reunir.unir.net/handle/123456789/19129 Self-Supervised Attentive Feature Learning for Alzheimer’s Disease Detection Elmannai, Hela; Saleem, Nasir; Bourouis, Sami; Alkanhel, Reem Ibrahim Alzheimer’s disease (AD) is a progressive neurodegenerative disorder that leads to memory loss and a decline in cognitive abilities. It primarily affects older adults and is the most common cause of dementia. Using deep learning, models can analyze brain imaging scans to detect specific patterns and biomarkers associated with the disease. Supervised learning models achieve high accuracy rates, but they require a large amount of data sets and labelled medical images. Self-supervised learning can achieve high accuracy rates with fewer training data. This study proposes a self-supervised attentive feature learning network (SSA-Net) for classifying Alzheimer’s disease. The proposed approach leverages self-supervised learning and attention mechanisms to enhance the accuracy and reliability of the classifying model. We employ ResNet-50, incorporating attentive activation, which replaces the ReLU activation, improving the ability of the neural model to focus on the most relevant features in the input medical images. We use SimCLR (Simple Framework for Contrastive Learning of Visual Representations) with the ResNet-50 backbone as a self-supervised learning framework that effectively learns high-quality visual representations in brain MRI (Magnetic Resonance Imaging) scans without labelling. We used the Kaggle Alzheimer’s classification dataset (KACD) containing brain MRI scans for training and testing. Experimental results on the KACD dataset show that the proposed attentive self-supervised ResNet50 reached 99.7% classification accuracy compared to the traditional ResNet50 with 98.1% accuracy. Evaluation metrics show the effectiveness of the proposed SSA-Net for the efficient classification of Alzheimer’s disease. Submitted by Angela María Porras Ruiz (angela.porras@unir.net) on 2026-03-06T10:42:32Z No. of bitstreams: 1 Self-Supervised Attentive Feature Learning for Alzheimer’s Disease Detection.pdf: 893051 bytes, checksum: 97b61d8ee0b13f3aea3f086ae205f8ea (MD5); Made available in DSpace on 2026-03-06T10:42:33Z (GMT). No. of bitstreams: 1 Self-Supervised Attentive Feature Learning for Alzheimer’s Disease Detection.pdf: 893051 bytes, checksum: 97b61d8ee0b13f3aea3f086ae205f8ea (MD5) https://reunir.unir.net/handle/123456789/19128 Attention Embedded Residual Bottleneck CNN Architecture for Breast Cancer Diagnosis in Ultrasound Images Fatima, Mamuna; Khan, Muhammad Attique; Shaheen, Saima; Kadry, Seifedine; Alqahtani, Omar; Alouane, M. Turki-Hadj Breast cancer (BrC) stands as the predominant cancer among women, resulting in a substantial global mortality toll each year. Early detection plays a pivotal role in diminishing mortality rates. Manual diagnosis of BrC is time-intensive, intricate, and prone to errors, emphasizing the necessity for an automated system for timely detection. Various imaging methods have been investigated, underscoring the crucial need for accurate detection to prevent unwarranted treatments and biopsies. Recent years have witnessed substantial exploration and enhancement in the application of DL for efficiently processing medical images. This study aiming to create an effective and resilient DL framework for BrC detection and classification. The steps are contrast enhancement and augmentation, a hybrid CNN network ‘BrC-DeepRBNet’ is introduced that is built from scratch and incorporates several design elements including residual blocks, bottleneck architecture, and a self-attention mechanism. This framework is employed to construct two networks, one comprising of 107 layers and the other with 149 layers. Moreover, the network capitalizes on the benefits offered by batch normalization (BN) and group normalization (GN), utilizes ReLU and leaky ReLU as activation functions, and integrates Max pooling layer into its architecture in a series of residual-bottleneck blocks. Further, for feature fusion horizontal approach is used and optimization is done using generalized normal distribution optimization (GNDO). The selected features are further classified using neural network classifiers. The introduced framework achieved the highest classification accuracy at 97.05% with publicly available BUS dataset. A detailed ablation study is presented that demonstrates the superior performance of the presented approach, surpassing various pre-trained models (i.e. AlexNet, InceptionV3, ResNet50, and ResNet101) and existing BrC detection and classification techniques. Submitted by Angela María Porras Ruiz (angela.porras@unir.net) on 2026-03-06T10:36:41Z No. of bitstreams: 1 Attention Embedded Residual Bottleneck CNN Architecture for Breast Cancer Diagnosis in Ultrasound Images.pdf: 1319325 bytes, checksum: 4835fb493fbddb081c3c59346e1bdd4c (MD5); Made available in DSpace on 2026-03-06T10:36:41Z (GMT). No. of bitstreams: 1 Attention Embedded Residual Bottleneck CNN Architecture for Breast Cancer Diagnosis in Ultrasound Images.pdf: 1319325 bytes, checksum: 4835fb493fbddb081c3c59346e1bdd4c (MD5) https://reunir.unir.net/handle/123456789/19127 Ensemble Diabetic Retinopathy Severity Classification Framework With Optimized VGG16, Resnet, and Inception Features Sheela, A. Jeba; Krishnamurthy, M. Background problem: Diabetic Retinopathy (DR) is characterized by high glucose levels in the blood, which can lead to permanent vision loss and microvascular complications. Various deep learning techniques for DR analysis tend to be more complex and may experience delays in delivering accurate results, thereby limiting their application in clinical settings. Implementing real-time predictionand severity analysisof DR can address this problem by providing real-time diagnostic insights based on DR severity levels. Aim: So, this paper is intended to offer a new DR detection and severity classification model with the highranking-based ensemble learning approach. Methodology: The preprocessed and segmented images are utilized in the feature extraction processusing ensemble architecture which incorporated VGG16, Resnet, and Inception to get three sets of features. The optimal features are selected using an Adaptive Scavenger-Based Dingo Optimization Algorithm (AS-DOX) to achieve the efficient classification of DR severity. The optimization constraint stake place in the HighRanking-Based Deep Ensemble Learning (HR-DEL) model helps to enhance the efficacy of classification for the offered approach. The simulation analysis provides enhanced performance with the accurate classification of the designed DR severity classification approach by comparing it with other baseline methods. Result: From the result analysis, the offered method achieves 96.6 % accuracy and sensitivity rate. Moreover, it achieves a 90.52% precision rate. Conclusion: Thus, the designed DR severity classification model attains better performance, and also it is utilized for early detection of DR severity. Submitted by Angela María Porras Ruiz (angela.porras@unir.net) on 2026-03-06T10:30:32Z No. of bitstreams: 1 Ensemble Diabetic Retinopathy Severity Classification Framework With Optimized VGG16, Resnet, and Inception Features.pdf: 1425672 bytes, checksum: 1c1ae01ff8659faf7cdf408c333e09ad (MD5); Made available in DSpace on 2026-03-06T10:30:32Z (GMT). No. of bitstreams: 1 Ensemble Diabetic Retinopathy Severity Classification Framework With Optimized VGG16, Resnet, and Inception Features.pdf: 1425672 bytes, checksum: 1c1ae01ff8659faf7cdf408c333e09ad (MD5) https://reunir.unir.net/handle/123456789/19115 On the Use of Large Language Models at Solving Math Problems: A Comparison Between GPT-4, LlaMA-2 and Gemini García Navarro, Alejandro L.; Koneva, Nataliia; Hernández, José Alberto; Sánchez-Macián, Alfonso In November 2022, ChatGPT v3.5 was announced to the world. Since then, Generative Artificial Intelligence (GAI) has appeared in the news almost daily, showing impressive capabilities at solving multiple tasks that have surprised the research community and the world in general. Indeed the number of tasks that ChatGPT and other Large Language Models (LLMs) can do are unimaginable, especially when dealing with natural text. Text generation, summarisation, translation, and transformation (into poems, songs, or other styles) are some of its strengths. However, when it comes to reasoning or mathematical calculations, ChatGPT finds difficulties. In this work, we compare different flavors of ChatGPT (v3.5, v4, and Wolfram GPT) at solving 20 mathematical tasks, from high school and first-year engineering courses. We show that GPT-4 is far more powerful than ChatGPT-3.5, and further that the use of Wolfram GPT can even slightly improve the results obtained with GPT-4 at these mathematical tasks. Submitted by Angela María Porras Ruiz (angela.porras@unir.net) on 2026-03-04T16:46:41Z No. of bitstreams: 1 On the Use of Large Language Models at Solving Math Problems.pdf: 416406 bytes, checksum: ec1a652ce78b2e3a3d30e6fb54f768e6 (MD5); Made available in DSpace on 2026-03-04T16:46:41Z (GMT). No. of bitstreams: 1 On the Use of Large Language Models at Solving Math Problems.pdf: 416406 bytes, checksum: ec1a652ce78b2e3a3d30e6fb54f768e6 (MD5) https://reunir.unir.net/handle/123456789/19081 Geometrics Assisted Rubbing Generation and Semantics Enhanced Detection for Small and Dense OBI Character Wan, Xiuan; Fang, Yuchun; Wu, Jiahua; Pan, Shouyong Character detection is essential for subsequent Oracle Bone Inscription (OBI) research. However, the lack of labeled data and the complexity of small and dense OBI characters are the main difficulties in OBI detection research. In this paper, we propose a framework for rubbing generation that can automatically build up largescale rubbing samples with verisimilar scenarios to noisy wild OBI through geometric and morphological construction combined with style transferring. Moreover, we propose a semantic-enhanced detection model aiming at small and dense OBI through the fusion of multi-resolution feature maps with the enriched feature in the YOLOv5s backbone. We introduce the higher resolution and the Soft-NMS into the proposed OBI detection model to solve the overlapping of small and dense OBI characters. The augmented dataset improves the performance of benchmark object detection models in the real OBI detection task when sufficient data is lacking. Furthermore, the proposed OBI detection model can provide easy and preferable access to OBI detection even with a small number of labeled data and obtain preferable results. Experiments ascertain the effectiveness of the proposed OBI generation framework and the proposed OBI detection model. Submitted by Angela María Porras Ruiz (angela.porras@unir.net) on 2026-02-25T16:34:26Z No. of bitstreams: 1 Geometrics Assisted Rubbing Generation and Semantics Enhanced Detection for Small and Dense OBI Character.pdf: 2212538 bytes, checksum: 09e6fef4d982a81652653756e6970bff (MD5); Made available in DSpace on 2026-02-25T16:34:26Z (GMT). No. of bitstreams: 1 Geometrics Assisted Rubbing Generation and Semantics Enhanced Detection for Small and Dense OBI Character.pdf: 2212538 bytes, checksum: 09e6fef4d982a81652653756e6970bff (MD5) https://reunir.unir.net/handle/123456789/19080 Three Dimensional Tree Modeling Based on the Skeleton Path Optimization and Geometrical Shapes Li, Xin; Zhou, Xuan; Xu, Sheng Nowadays, the 3D individual tree reconstruction has played a significant role in the phenotypic study of trees. This paper proposes a new automatic method for extracting skeletons of individual trees and reconstructing 3D models. Firstly, the Euclidean clustering is performed to obtain center points of candidate branch regions. Then, the initial skeletons of LiDAR point clouds are obtained by slicing clusters in three dimensions. Secondly, skeleton points are completed by the proposed branch tracking. Then, the radius of the branches is accurately estimated from the branches. Thirdly, optimal points are interpolated in appropriate directions to refine skeletons of individual trees. Then, the Laplacian algorithm is conducted for smoothing branches. After that, optimal geometric shapes are formulated to reconstruct the final 3D tree models. Experimental results show that the average accuracy of our individual tree models is up to 97.49%, which shows a promising algorithm in 3D tree reconstructions. Submitted by Angela María Porras Ruiz (angela.porras@unir.net) on 2026-02-25T16:27:48Z No. of bitstreams: 1 Three Dimensional Tree Modeling Based on the Skeleton Path Optimization and Geometrical Shapes.pdf: 1783238 bytes, checksum: 78e5941b5dd4a843db7d0c416d3cf06d (MD5); Made available in DSpace on 2026-02-25T16:27:48Z (GMT). No. of bitstreams: 1 Three Dimensional Tree Modeling Based on the Skeleton Path Optimization and Geometrical Shapes.pdf: 1783238 bytes, checksum: 78e5941b5dd4a843db7d0c416d3cf06d (MD5) https://reunir.unir.net/handle/123456789/19079 Posture Estimation of Curve Running Motion Using Nano-Biosensor and Machine Learning Wu, Xiaoming; Cao, Yu; Wang, Yu; Li, Bing; Yang, Haitao; Raja, S. P. Curve running is a common form of training and competition. Conducting research on posture estimation during curve running can provide more accurate training and competition data for athletes. However, due to the unique nature of curve running, traditional posture estimation methods neglect the temporal changes in athlete posture, resulting in a decrease in estimation accuracy. Therefore, a posture estimation method for curve running motion using nano-biosensor and machine learning is proposed. First, the motion parameters of humans are collected by nano-biosensor, and the posture coordinates are obtained preliminarily. Second, the posture coordinates are established according to the human motion parameters, and the curve running posture data is obtained and filtered to obtain more accurate data. Finally, the Bayesian network in machine learning is used to continuously track the posture, and a nonlinear equation is established to fuse the posture angle obtained by the sensor and the posture tracked by the Bayesian network, to realize the posture estimation of curve running motion. The results show that the proposed estimation method has a good motion posture estimation effect, and the hip joint estimation error, knee joint estimation error and ankle joint estimation error are all less than 5°, and the endpoint displacement estimation offset rate is less than 2%. It can realize accurate motion posture estimation of curve running motion, and has important application value in the field of track training. Submitted by Angela María Porras Ruiz (angela.porras@unir.net) on 2026-02-25T16:21:22Z No. of bitstreams: 1 Posture Estimation of Curve Running Motion Using Nano Biosensor and Machine Learning.pdf: 508410 bytes, checksum: 9f8033f27a6485bc48c75b4c6ca909f9 (MD5); Made available in DSpace on 2026-02-25T16:21:22Z (GMT). No. of bitstreams: 1 Posture Estimation of Curve Running Motion Using Nano Biosensor and Machine Learning.pdf: 508410 bytes, checksum: 9f8033f27a6485bc48c75b4c6ca909f9 (MD5) https://reunir.unir.net/handle/123456789/19078 A Realtime Classroom Assessment System for Analysis of Students’ Evaluation of Teaching Through a Deep Learning and Emotional Contagion Mechanism Lin, Kuan-Cheng; Lin, Ya-Hsuan; Chen, Ya-Hsuan Student evaluations of teacher performance are often derived from end-of-semester assessments, significantly impacting the authenticity of teaching evaluations but failing to provide real-time feedback. In addition, teachers' emotional states affect student performance, including in terms of learning motivation and classroom participation, which reflect the students' emotional state. This teacher-student emotional contagion mechanism focuses on the interaction of teacher-student emotions and can be used to observe the quality of instructional performance. Therefore, automatically detecting teacher-student emotional interaction and then providing real-time class satisfaction feedback can provide teachers with a more effective basis for adjusting classroom content. This research proposes an end-to-end classroom real-time teaching evaluation system based on automatic facial-emotion recognition, which can accurately detect and directly analyze the emotions of students and teachers in streaming frames. The system consists of two parts: First, a YOLO model based on deep learning approaches is used to automatically detect the emotional states of teachers and students during the teaching process; Then, combining the emotional contagion mechanism with the teaching evaluation scale, teaching satisfaction can be predicted using a Long Short-Term Memory (LSTM) model to output a classroom satisfaction score within a fixed period. Further analysis of the testing dataset confirms that the model has a high reliability in predicting teaching satisfaction. Research results show the proposed system can achieve an emotional recognition accuracy rate of 98.1% for teachers and 99.5% for students based on the emotion datasets. Further development could potentially provide teachers with strategies to improve classroom teaching effectiveness, better understand students' emotions and learning motivation, and improve learning outcomes. Submitted by Angela María Porras Ruiz (angela.porras@unir.net) on 2026-02-25T16:15:50Z No. of bitstreams: 1 A Realtime Classroom Assessment System for Analysis.pdf: 410131 bytes, checksum: 7244389b1751d418174f081e546ef794 (MD5); Made available in DSpace on 2026-02-25T16:15:50Z (GMT). No. of bitstreams: 1 A Realtime Classroom Assessment System for Analysis.pdf: 410131 bytes, checksum: 7244389b1751d418174f081e546ef794 (MD5) https://reunir.unir.net/handle/123456789/19077 ChatGPT, Generative AI, Mathematical Problems, Wolfram Mathematica García Navarro, Alejandro L.; Koneva, Nataliia; Hernández, José Alberto; Sánchez-Macián, Alfonso In November 2022, ChatGPT v3.5 was announced to the world. Since then, Generative Artificial Intelligence (GAI) has appeared in the news almost daily, showing impressive capabilities at solving multiple tasks that have surprised the research community and the world in general. Indeed the number of tasks that ChatGPT and other Large Language Models (LLMs) can do are unimaginable, especially when dealing with natural text. Text generation, summarisation, translation, and transformation (into poems, songs, or other styles) are some of its strengths. However, when it comes to reasoning or mathematical calculations, ChatGPT finds difficulties. In this work, we compare different flavors of ChatGPT (v3.5, v4, and Wolfram GPT) at solving 20 mathematical tasks, from high school and first-year engineering courses. We show that GPT-4 is far more powerful than ChatGPT-3.5, and further that the use of Wolfram GPT can even slightly improve the results obtained with GPT-4 at these mathematical tasks. Submitted by Angela María Porras Ruiz (angela.porras@unir.net) on 2026-02-25T16:07:52Z No. of bitstreams: 1 On the Use of Large Language Models at Solving Math Problems.pdf: 416406 bytes, checksum: ec1a652ce78b2e3a3d30e6fb54f768e6 (MD5); Made available in DSpace on 2026-02-25T16:07:52Z (GMT). No. of bitstreams: 1 On the Use of Large Language Models at Solving Math Problems.pdf: 416406 bytes, checksum: ec1a652ce78b2e3a3d30e6fb54f768e6 (MD5) https://reunir.unir.net/handle/123456789/19076 A Sustainable Deep Learning Paradigm for Reliable Energy Prediction in Next-Gen Consumer Electronics Jeon, Gwangil; Ahmed, Imran; Han, Sanghoon In the rapidly evolving consumer electronics landscape, the imperative for sustainable energy solutions necessitates the development of accurate energy prediction methodologies. Traditional energy prediction models often fall short in accounting for the dynamic characteristics of renewable energy sources, particularly wind and solar. This limitation is pronounced in consumer electronics, where precise energy forecasting is pivotal for achieving optimal device performance and energy efficiency. To address this gap, we present a sustainable deep learning paradigm using Long Short-Term Memory (LSTM) networks to capture the complex temporal patterns inherent in renewable energy data. This paper introduces a novel and sustainable deep learning approach that significantly enhances energy prediction accuracy within the context of next-generation consumer electronics. By leveraging the capabilities of an LSTM-based model, we utilize an extensive dataset comprising hourly records of wind and solar energy production from the French grid since 2020. Our research addresses the inherent challenges in precise energy prediction, a cornerstone for efficient energy management and consumption optimization in emerging technology ecosystems. Through comprehensive data preprocessing, feature engineering, and rigorous training, the LSTM model demonstrates exceptional proficiency, achieving an impressive 82% accuracy in predicting energy production. This underscores its efficacy in capturing intricate temporal relationships and patterns within renewable energy data, facilitating its integration into next-generation consumer electronics. Our proposed paradigm addresses a critical need and paves the way for a future where accurate energy prediction fuels eco-conscious technology. In conclusion, this study contributes to a more sustainable energy landscape by advancing the development of reliable and efficient energy prediction methodologies for the evolving realm of next-generation consumer electronics. Submitted by Angela María Porras Ruiz (angela.porras@unir.net) on 2026-02-25T16:01:01Z No. of bitstreams: 1 A Sustainable Deep Learning Paradigm for Reliable Energy Prediction in Next Gen Consumer Electronics.pdf: 572480 bytes, checksum: 15dbecabf4998c3cebec7cfc3700902d (MD5); Made available in DSpace on 2026-02-25T16:01:01Z (GMT). No. of bitstreams: 1 A Sustainable Deep Learning Paradigm for Reliable Energy Prediction in Next Gen Consumer Electronics.pdf: 572480 bytes, checksum: 15dbecabf4998c3cebec7cfc3700902d (MD5) https://reunir.unir.net/handle/123456789/19075 IAtraj: Multi-Modal Trajectory Prediction Through Contextual Information Spatio-Temporal Interaction and Awareness Wang, Xiaoliang; Zhou, Lian; Li, Kuan-Ching; Zheng, Shiqi; Fan, Huijing Accurately and feasibly predicting the future trajectories of autonomous vehicles is a critically important task. However, this task faces significant challenges due to the variability of driving intentions and the complexity of social interactions. These challenges primarily arise from the need to understand one’s driving behaviors and model the interaction information of the surrounding environment. A substantial amount of research has been focused on integrating interaction information from the surrounding environment, mainly using raster images or High-Definition maps (HD maps). However, the real-time update of environmental maps and the high computational cost associated with processing interaction information using compatible technologies such as vision have become limiting factors. Additionally, ineffective simulation and modeling of real driving scenarios, coupled with inadequate understanding of contextual environmental information, result in lower prediction accuracy. To overcome these challenges, we propose a multi-modal trajectory prediction model based on sequence modeling namely IAtraj, incorporating multiple attention mechanisms, focuses on the three critical elements in real traffic scenarios: the target agent’s historical trajectory, effective interactions with neighboring vehicles, and lane supervision and retention strategies. To better model these elements, we design modules for Temporal Interaction (TI), Spatial Interaction (SI), and Lane Awareness (LA). Through extensive experiments conducted on the publicly available nuScenes dataset, IAtraj exhibits outstanding performance, successfully addressing the challenges of temporal dependencies in trajectory sequences and the representation of scene changes. Finally, comprehensive ablation experiments validate the effectiveness of each significant module, reinforcing the reliability and robustness of IAtraj in dealing with complex traffic scenarios. Submitted by Angela María Porras Ruiz (angela.porras@unir.net) on 2026-02-25T15:46:42Z No. of bitstreams: 1 IAtraj Multi-Modal Trajectory Prediction Through Contextual Information Spatio-Temporal Interaction and Awareness.pdf: 1423128 bytes, checksum: 60d5f9e813b3a6ff8100670b5134b89d (MD5); Made available in DSpace on 2026-02-25T15:46:42Z (GMT). No. of bitstreams: 1 IAtraj Multi-Modal Trajectory Prediction Through Contextual Information Spatio-Temporal Interaction and Awareness.pdf: 1423128 bytes, checksum: 60d5f9e813b3a6ff8100670b5134b89d (MD5) https://reunir.unir.net/handle/123456789/19073 A 2D Clustering Based Hotspot Identification Approach for Spatio-Temporal Crime Prediction Iqbal, Muhammad Faisal Buland; Ullah, Aman; Alhomoud, Ahmed; Hussain, Tariq; Attar, Razaz Waheeb; Ouyang, Jianquan; Alnfiai, Mrim M.; Hatamleh, Wesam Atef This research introduces a method for predicting where crimes will occur based on clustering activity in the area. Hotspots, or locations with a disproportionately high number of crimes, are located by a combination of spatial and temporal grouping methods employed by this strategy. Crime forecasting models use these hotspots to predict where crimes will occur. The approach's efficacy is tested using actual crime data, and it successfully predicts future crimes in high-crime zones. Law enforcement agencies can use the proposed method to protect the public better, and it shows promise as a tool for crime prediction. Academic research into the topic of foreseeing criminal behavior is a newer development. Researchers in this discipline have discovered that criminal behavior has universal patterns. These patterns may help law enforcement agencies plan for criminal activities. Predictive policing, hotspot analysis, and geographical profiling are examples of when crime forecasting has been useful. Several aspects of the census, such as the average yearly income and literacy rate, are related to the prevalence of crime in a certain area. Indicators of potentially criminal behavior, these characteristics may be seen as markers. This investigation aims to discover if any clues can be gleaned from past criminal behavior that may be utilized to forecast future criminal behavior. Using machine learning and 2-D Hotspot analysis, we propose a method for the spatiotemporal prediction of criminal activity within the scope of this study. Clustering is a method used in 2-dimensional hotspot analysis. Methods of modern categorization, both with and without hotspot analysis, are used to evaluate the suggested model's efficacy. It is found that the model that incorporates hotspot analysis performs better than the one that does not. Submitted by Angela María Porras Ruiz (angela.porras@unir.net) on 2026-02-25T15:27:47Z No. of bitstreams: 1 A 2D Clustering Based Hotspot Identification Approach for Spatio-Temporal Crime Prediction.pdf: 561649 bytes, checksum: 3df958bf45fed378586ac4711cc20fc4 (MD5); Made available in DSpace on 2026-02-25T15:27:47Z (GMT). No. of bitstreams: 1 A 2D Clustering Based Hotspot Identification Approach for Spatio-Temporal Crime Prediction.pdf: 561649 bytes, checksum: 3df958bf45fed378586ac4711cc20fc4 (MD5)