Information

16 pages, 1996 KiB

Open AccessArticle

Movement Compensation in Dual Continuous Wave Radar Using Deep Learning

by Gonçalo Gomes, Susana Brás, Carolina Gouveia, Daniel Albuquerque and Pedro Pinho

Information 2025, 16(2), 99; https://doi.org/10.3390/info16020099 - 31 Jan 2025

This work presents an innovative method for detecting the respiratory patterns of subjects walking on a treadmill, by leveraging the capabilities of deep learning (DL) technologies and a dual-radar setup. The study aims to overcome the challenge of accurately capturing respiratory rates in [...] Read more.

This work presents an innovative method for detecting the respiratory patterns of subjects walking on a treadmill, by leveraging the capabilities of deep learning (DL) technologies and a dual-radar setup. The study aims to overcome the challenge of accurately capturing respiratory rates in subjects performing body movements, a scenario less addressed in prior studies. By employing two radars operating at

5.8

GHz for motion mitigation, this study compares the efficacy of dual-radar configurations against a single-radar setup. The study employs DL algorithms based on convolutional autoencoders to mitigate the low-quality demodulated radar signals by reconstructing the respiratory signal. The models are trained with data from a single subject and data from 15 subjects, attaining average absolute errors of

0.29

and

4.59

Respiration Per Minute (RPM), respectively, allowing to conclude that the use of DL algorithms enhances the accuracy of respiratory signal detection when compared with arctangent demodulation, even in cases where radar data contain minimal information regarding vital signals. Full article

(This article belongs to the Section Information Applications)

19 pages, 6648 KiB

Open AccessArticle

Research on Resource Allocation Algorithm for Non-Orthogonal Multiple Access Backscatter-Based Cognitive Radio Networks

by Tingpei Huang, Tiantian Zhang, Bairen Zhang, Jianhang Liu and Shibao Li

Information 2025, 16(2), 98; https://doi.org/10.3390/info16020098 - 31 Jan 2025

Abstract

Integrating Non-Orthogonal Multiple Access (NOMA) backscatter with Cognitive Radio (CR) can effectively alleviate the pressure of large-scale Internet of Things (IoT) device connections while reducing power consumption. In the downlink of NOMA backscatter-based CR networks (NB-CR), most researchers only consider the case of [...] Read more.

Integrating Non-Orthogonal Multiple Access (NOMA) backscatter with Cognitive Radio (CR) can effectively alleviate the pressure of large-scale Internet of Things (IoT) device connections while reducing power consumption. In the downlink of NOMA backscatter-based CR networks (NB-CR), most researchers only consider the case of a single backscatter device (BD), ignoring the fact that multiple BDs can prolong device usage and enhance system robustness. The resource allocation (RA) problem is crucial in the downlink of NB-CR networks. Most existing RA algorithms focus on system throughput but rarely consider energy efficiency (EE). In this paper, we propose the RA problem for downlink communication (NBCR-RA) in NB-CR networks with multiple BDs. We jointly optimize power allocation coefficients and reflection coefficients (RCs) to maximize EE. We model the NBCR-RA problem as a non-convex problem and divide it into two subproblems: power allocation optimization and RC optimization. Firstly, we propose a Lagrange-based power allocation optimization (L-PA) algorithm to obtain the optimal power allocation coefficients. Secondly, we design an RC optimization algorithm, PS-RC, based on a particle swarm algorithm to determine the optimal RCs. Finally, we validate the superiority of L-PA and PS-RC algorithms in terms of EE. Through multiple experiments, we obtained a

95 %

confidence interval of [10.5552, 10.6465]. Full article

(This article belongs to the Section Wireless Technologies)

► Show Figures

Graphical abstract

24 pages, 799 KiB

Open AccessArticle

Unveiling XSS Threats: A Bipartite Graph Approach with Ensemble Deep Learning for Enhanced Detection

by Wafa Alorainy

Information 2025, 16(2), 97; https://doi.org/10.3390/info16020097 - 31 Jan 2025

Abstract

Cross-Site Scripting (XSS) attacks are a common source of vulnerability for web applications, necessitating scalable mechanisms for detection. In this work, a new method based on bipartite graph-based feature extraction and an ensemble learning classifier containing CNN, LSTM, and GRU is introduced. Our [...] Read more.

Cross-Site Scripting (XSS) attacks are a common source of vulnerability for web applications, necessitating scalable mechanisms for detection. In this work, a new method based on bipartite graph-based feature extraction and an ensemble learning classifier containing CNN, LSTM, and GRU is introduced. Our proposed bipartite graph model is novel as the payloads constitute the first set, while the words constructing the payloads comprise the second set. This representation allows structural and contextual dependencies to be extracted so the model can recognize complex and obfuscated XSS payloads. Our method surpasses state-of-the-art methods by having 99.97% detection accuracy. It has a significantly increased ability to detect complicated payload variations by utilizing co-occurrence patterns and interdependence between smaller payload parts through the adoption of these bipartite features. In addition to improving the F1-score, recall, and precision associated with such methods, it also demonstrates the adaptability of graph-based representation in cybersecurity applications. Our findings highlight the possibility of integrating ensemble classifiers and refined feature engineering into a scalable, precise XSS detection system. Full article

23 pages, 4237 KiB

Open AccessArticle

Fostering STEM Skills Through Programming and Robotics for Motivation and Cognitive Development in Secondary Education

by Iván Torres and Esteban Inga

Information 2025, 16(2), 96; https://doi.org/10.3390/info16020096 - 31 Jan 2025

Abstract

This paper seeks to identify the impact of learning programming and robotics in the Science, Technology, Engineering, and Mathematics (STEM) educational approach. Studying these areas of knowledge is important to prepare students to face contemporary technological challenges. The approach analyzes how to establish [...] Read more.

This paper seeks to identify the impact of learning programming and robotics in the Science, Technology, Engineering, and Mathematics (STEM) educational approach. Studying these areas of knowledge is important to prepare students to face contemporary technological challenges. The approach analyzes how to establish and define the curricular content articulated in developing critical 21st-century skills within the teaching–learning process. A methodological strategy is proposed in the scientific field using the historical-descriptive method to carry out a literature review and a bibliometric study, evaluating scientific articles indexed in Web of Science (WoS) and Scopus from 2020 to 2024. Later, an evaluation is carried out using satisfaction surveys directed to eighth-grade students and teachers of the Unidad Educativa Fiscal Ciudad de Girón. These surveys address various aspects related to the context of learning programming and robotics from the STEM perspective. Consequently, the analytic–synthetic approach revealed that teaching programming and robotics would promote cognitive skills from adolescence, which is crucial for building solid foundations in STEM concepts. The positive impact on the motivation for change in students and teachers is highlighted by facilitating interaction with technologies and applying knowledge in practical projects in the educational process. Full article

37 pages, 1220 KiB

Open AccessReview

Shaping the Future of Higher Education: A Technology Usage Study on Generative AI Innovations

by Weina Pang and Zhe Wei

Information 2025, 16(2), 95; https://doi.org/10.3390/info16020095 - 27 Jan 2025

Abstract

Generative Artificial Intelligence (GAI) is rapidly reshaping the landscape of higher education, offering innovative solutions to enhance student engagement, personalize learning experiences, and improve academic performance prediction. This study provides an in-depth exploration of GAI applications in educational contexts, drawing insights from 67 [...] Read more.

Generative Artificial Intelligence (GAI) is rapidly reshaping the landscape of higher education, offering innovative solutions to enhance student engagement, personalize learning experiences, and improve academic performance prediction. This study provides an in-depth exploration of GAI applications in educational contexts, drawing insights from 67 case studies meticulously selected from over 300 papers presented at the AIED 2024 conference. The research focuses on eight key themes from student engagement and behavior analysis to the integration of generative models into educational tools. These case studies illustrate the potential of GAI to optimize teaching practices, enhance student support systems, and provide tailored interventions that address individual learning needs. However, this study also highlights challenges such as scalability, the need for balanced and diverse datasets, and ethical concerns regarding data privacy and bias. Further, it emphasizes the importance of improving model accuracy, transparency, and real-world applicability in educational settings. The findings underscore the need for continued research to refine GAI technologies, ensuring they are scalable, adaptable, and equitable, ultimately enhancing the effectiveness and inclusivity of AI-driven educational tools across diverse higher education environments. It should be noted that this study primarily draws from papers presented at the AIED 2024 conference, which may limit global representativeness and introduce thematic biases. Future studies are encouraged to include broader datasets from diverse conferences and journals to ensure a more comprehensive understanding of GAI applications in higher education. Full article

(This article belongs to the Special Issue Generative AI Technologies: Shaping the Future of Higher Education)

35 pages, 954 KiB

Open AccessArticle

Charging Method Selection of a Public Charging Station Using an Interval-Valued Picture Fuzzy Bidirectional Projection Based on VIKOR Method with Unknown Attribute Weights

by Chittaranjan Shit and Ganesh Ghorai

Information 2025, 16(2), 94; https://doi.org/10.3390/info16020094 - 26 Jan 2025

Abstract

Excessive use of fossil fuel-powered vehicles is a major problem for the entire world today, because of which greenhouse gases are increasing day by day. As a result, climate change and global warming have grown to be serious problems that affect both the [...] Read more.

Excessive use of fossil fuel-powered vehicles is a major problem for the entire world today, because of which greenhouse gases are increasing day by day. As a result, climate change and global warming have grown to be serious problems that affect both the environment and life on Earth. However, the effective way of reducing greenhouse gases is to use electric vehicles for commuting. The assessment and selection of the best possible way of charging an electric vehicle is a convoluted decision-making challenge due to the presence of assorted contradictory criteria. Additionally, individual decision makers’ minds and insufficient data are obstacles to doing this. In this regard, interval-valued picture fuzzy sets have been considered as a compatible tool to handle vagueness. In this paper, a multi-attribute group decision-making problem with the bidirectional projection-based VlseKriterijumska Optimizacija I Kompromisno Resenje

(V I K O R)

method is considered where the weights are partially known. The objective weights of the attributes in this model are determined using the deviation-based approach. The compromised solution is also assessed using the

V I K O R

approach. Both the interval-valued image fuzzy Schweizer–Sklar power weighted geometric operator and the interval-valued picture fuzzy Schweizer–Sklar power weighted averaging operator are used in this process. Lastly, a numerical example showing the most suitable way to charge an electric vehicle is given to demonstrate the suggested methodology. To evaluate the robustness and efficacy of the suggested strategy, a comparative analysis with current techniques and a sensitivity analysis of the parameters are also carried out. Full article

(This article belongs to the Special Issue New Trend on Fuzzy Systems and Intelligent Decision Making Theory: A Themed Issue Dedicated to Dr. Ronald R. Yager)

► Show Figures

Figure 1

35 pages, 3825 KiB

Open AccessArticle

An Intelligent Model for Parametric Cognitive Assessment of E-Learning-Based Students

by Muhammad Saqib Javed, Muhammad Aslam and Syed Khaldoon Khurshid

Information 2025, 16(2), 93; https://doi.org/10.3390/info16020093 - 26 Jan 2025

Abstract

In an e-learning environment, question levels are based on Bloom’s Taxonomy (BT), which normally classifies a course’s learning objectives into diverse levels. As per the previous literature, the assessment procedure lacks accuracy and results in redundant keywords when automatically assigning Bloom’s taxonomic categories [...] Read more.

In an e-learning environment, question levels are based on Bloom’s Taxonomy (BT), which normally classifies a course’s learning objectives into diverse levels. As per the previous literature, the assessment procedure lacks accuracy and results in redundant keywords when automatically assigning Bloom’s taxonomic categories using a keyword-based approach. These assessments are considered challenging as far as e-learning-based students are concerned, as the text feed is the only instrumental testing part. Student assessments are limited to multiple-choice questions and lack an evaluation of students’ text-based input. This paper proposes a natural-language processing-based intelligent deep-learning model that relies on parametric cognitive assessments. By applying class labels to students’ descriptive responses, the proposed approach helps classify a variety of questions mapped to BT levels. The first contribution of this work is a compiled dataset of the assessment items from 300 students, who were tested on 20 questions at each level. Each level is calculated by combining the responses from all students, resulting in 6000 questions per cognitive level for a total of 36,000 records. The second contribution is the development of an intelligent model based on a recurrent neural network (RNN), which not only predicts Bloom’s question level but also learns it over further iterations. The students’ text-based answers are accessed to gauge performance using a refined question pool gathered through the RNN model. The student dataset is mapped and tested using the NLP model for further classification of the students’ cognitive levels. This assessment is related to the formulation of questions and the compilation of Episode 2 for assessment. The third contribution is the comparison and demonstration of the improvements in learning using a parametric cognitive-based assessment in an episodic manner. Improved classification accuracy was attained by adding more processing layers based on the iterative, RNN-based learning model to achieve the vital threshold difference. The cognitive based questions pool classification achieved by RNN results in 98% accuracy. The resulting student assessments, based on performance, increased to an accuracy ratio of 92.16% and a precision ratio of 92.36% at an aggregate level based on the Random Forest classifier. We claim that our work serves as an initiative for effective student evaluations in interactive and e-learning-based environments when handling other types of inputs, like mathematical, graphical, and multimodal inputs. Full article

(This article belongs to the Special Issue Intelligent Agent and Multi-Agent System)

► Show Figures

Graphical abstract

21 pages, 2056 KiB

Open AccessArticle

EquiFlowShard: A Blockchain Sharding Protocol with Optimized Account Distribution

by Yemin Chen, Yongdong Wu and Tong Li

Information 2025, 16(2), 92; https://doi.org/10.3390/info16020092 - 25 Jan 2025

Abstract

Blockchain sharding is a scalable solution for distributed ledgers, but may be hindered due to cross-shard transactions and uneven workload distribution. This paper presents EquiFlowShard, an advanced blockchain sharding protocol designed to improve robustness and enhance cross-shard efficiency. Specifically, by employing Optimized Account [...] Read more.

Blockchain sharding is a scalable solution for distributed ledgers, but may be hindered due to cross-shard transactions and uneven workload distribution. This paper presents EquiFlowShard, an advanced blockchain sharding protocol designed to improve robustness and enhance cross-shard efficiency. Specifically, by employing Optimized Account State Distribution Algorithm (OSADA), EquiFlowShard dynamically assigns and segments account states, so as to minimize cross-shard transaction volume and balance shard workloads. In addition, the protocol introduces the SFlow mechanism to facilitate secure and consistent state transfers and a Smooth Transition scheme to mitigate performance impacts during state reconfigurations. Evaluation results confirm that EquiFlowShard outperforms existing benchmark protocols in terms of throughput, transaction confirmation latency, and cross-shard transaction ratio, demonstrating its effectiveness in dynamic blockchain environments. Full article

17 pages, 4505 KiB

Open AccessArticle

An Application of SEMAR IoT Application Server Platform to Drone-Based Wall Inspection System Using AI Model

by Yohanes Yohanie Fridelin Panduman, Radhiatul Husna, Noprianto, Nobuo Funabiki, Shunya Sakamaki, Sritrusta Sukaridhoto, Yan Watequlis Syaifudin and Alfiandi Aulia Rahmadani

Information 2025, 16(2), 91; https://doi.org/10.3390/info16020091 - 24 Jan 2025

Abstract

Recently, artificial intelligence (AI) has been adopted in a number of Internet of Things (IoT) application systems to enhance intelligence. We have developed a ready-made server with rich built-in functions to collect, process, display, analyze, and store data from various IoT devices, the [...] Read more.

Recently, artificial intelligence (AI) has been adopted in a number of Internet of Things (IoT) application systems to enhance intelligence. We have developed a ready-made server with rich built-in functions to collect, process, display, analyze, and store data from various IoT devices, the SEMAR (Smart Environmental Monitoring and Analytics in Real-Time) IoT application server platform, in which various AI techniques have been implemented to enhance its capabilities. In this paper, we present an application of SEMAR to a drone-based wall inspection system using an object detection AI model called You Only Look Once (YOLO). This system aims to detect wall cracks at high places using images taken via a camera on a flying drone. An edge computing device is installed to control the drone, sending the taken images through the Kafka system, storing them with the drone flight data, and sending the data to SEMAR. The images are analyzed via YOLO through SEMAR. For evaluations, we implemented the system using Ryze Tello for the drone and Raspberry Pi for the edge, and we evaluated the detection accuracy. The preliminary experiment results confirmed the effectiveness of the proposal. Full article

(This article belongs to the Special Issue Advances in Machine Learning and Intelligent Information Systems)

► Show Figures

Graphical abstract

17 pages, 1985 KiB

Open AccessArticle

A Spine-Specific Lexicon for the Sentiment Analysis of Interviews with Adult Spinal Deformity Patients Correlates with SF-36, SRS-22, and ODI Scores: A Pilot Study of 25 Patients

by Ross Gore, Michael M. Safaee, Christopher J. Lynch and Christopher P. Ames

Information 2025, 16(2), 90; https://doi.org/10.3390/info16020090 - 24 Jan 2025

Abstract

Classic health-related quality of life (HRQOL) metrics are cumbersome, time-intensive, and subject to biases based on the patient’s native language, educational level, and cultural values. Natural language processing (NLP) converts text into quantitative metrics. Sentiment analysis enables subject matter experts to construct domain-specific [...] Read more.

Classic health-related quality of life (HRQOL) metrics are cumbersome, time-intensive, and subject to biases based on the patient’s native language, educational level, and cultural values. Natural language processing (NLP) converts text into quantitative metrics. Sentiment analysis enables subject matter experts to construct domain-specific lexicons that assign a value of either negative (−1) or positive (1) to certain words. The growth of telehealth provides opportunities to apply sentiment analysis to transcripts of adult spinal deformity patients’ visits to derive a novel and less biased HRQOL metric. In this study, we demonstrate the feasibility of constructing a spine-specific lexicon for sentiment analysis to derive an HRQOL metric for adult spinal deformity patients from their preoperative telehealth visit transcripts. We asked each of twenty-five (25) adult patients seven open-ended questions about their spinal conditions, treatment, and quality of life during telehealth visits. We analyzed the Pearson correlation between our sentiment analysis HRQOL metric and established HRQOL metrics (the Scoliosis Research Society-22 questionnaire [SRS-22], 36-Item Short Form Health Survey [SF-36], and Oswestry Disability Index [ODI]). The results show statistically significant correlations (0.43–0.74) between our sentiment analysis metric and the conventional metrics. This provides evidence that applying NLP techniques to patient transcripts can yield an effective HRQOL metric. Full article

(This article belongs to the Special Issue Biomedical Natural Language Processing and Text Mining)

► Show Figures

Figure 1

20 pages, 1353 KiB

Open AccessArticle

Discourse Analysis of International Scientific Organizations in the 2022 Russia–Ukraine Conflict: A Natural Language Processing Approach

by Jiayue Lu, Xiaoli Chen and Xuezhao Wang

Information 2025, 16(2), 89; https://doi.org/10.3390/info16020089 - 24 Jan 2025

Abstract

The scientific community has not stayed outside the Russia–Ukraine conflict. This study analyzes the attitudes and roles of international scientific organizations in the conflict, based on 923 official statements, through a combination of discourse analysis and Natural Language Processing (NLP) techniques, including sentiment [...] Read more.

The scientific community has not stayed outside the Russia–Ukraine conflict. This study analyzes the attitudes and roles of international scientific organizations in the conflict, based on 923 official statements, through a combination of discourse analysis and Natural Language Processing (NLP) techniques, including sentiment analysis and topic modeling. The findings reveal that 527 organizations issued statements, with 47% explicitly “supporting Ukraine and condemning Russia”, and 13% maintaining a neutral stance. These statements reflect diverse concerns, including the conflict’s immediate humanitarian impact, disruption to scientific collaboration, and broader political and social implications. This research contributes to understanding how international scientific organizations navigate conflict contexts by systematically uncovering their attitudes, focus areas, and actions. Through a thematic analysis, the study demonstrates how these organizations articulate their positions, advocate for specific measures, and leverage their influence to address issues such as economic support, scientific collaboration, and healthcare assistance. By identifying these behaviors, the study clarifies the strategic roles scientific organizations play in shaping discourse and mediating international relations, offering key insights into their impact during geopolitical crises. Full article

(This article belongs to the Special Issue 2nd Edition of Information Retrieval and Social Media Mining)

► Show Figures

Figure 1

22 pages, 2493 KiB

Open AccessArticle

A Data-Driven Approach to Estimating the Impact of Audiovisual Art Events Through Web Presence

by Andreas Giannakoulopoulos, Minas Pergantis, Aristeidis Lamprogeorgos and Stella Lampoura

Information 2025, 16(2), 88; https://doi.org/10.3390/info16020088 - 24 Jan 2025

Abstract

In the modern era, a major avenue of dissemination for cultural and artistic events is through the World-Wide Web, where every such event has a multifaceted distinct digital footprint. This digital footprint is an indicator of how strong the influence of each event [...] Read more.

In the modern era, a major avenue of dissemination for cultural and artistic events is through the World-Wide Web, where every such event has a multifaceted distinct digital footprint. This digital footprint is an indicator of how strong the influence of each event is in the public’s perception and to what extent it becomes part of the audiovisual art landscape. This study aims to present how the impact of an audiovisual event may be estimated using quantitative data collected through its online presence. This data-driven approach is made possible through web data extraction techniques and the use of generative artificial intelligence, which allows for structured information extraction from an endless variety of websites. Based on an event’s innate characteristics, web outreach, estimated scope, and thematical popularity, an encompassing impact factor is calculated, which may be used to rank events on the basis of perceived influence. For the purposes of this study, a dataset consisting of thousands of events in Greece was collected over an extended period. These data were used for a computational statistical analysis. Through this process of data collection, impact calculation, and analysis, data-driven insights were derived concerning the landscape of audiovisual art events. Full article

(This article belongs to the Section Information Applications)

► Show Figures

Figure 1

36 pages, 1347 KiB

Open AccessReview

Transitioning from MLOps to LLMOps: Navigating the Unique Challenges of Large Language Models

by Saurabh Pahune and Zahid Akhtar

Information 2025, 16(2), 87; https://doi.org/10.3390/info16020087 - 23 Jan 2025

Abstract

Large Language Models (LLMs), such as the GPT series, LLaMA, and BERT, possess incredible capabilities in human-like text generation and understanding across diverse domains, which have revolutionized artificial intelligence applications. However, their operational complexity necessitates a specialized framework known as LLMOps (Large Language [...] Read more.

Large Language Models (LLMs), such as the GPT series, LLaMA, and BERT, possess incredible capabilities in human-like text generation and understanding across diverse domains, which have revolutionized artificial intelligence applications. However, their operational complexity necessitates a specialized framework known as LLMOps (Large Language Model Operations), which refers to the practices and tools used to manage lifecycle processes, including model fine-tuning, deployment, and LLMs monitoring. LLMOps is a subcategory of the broader concept of MLOps (Machine Learning Operations), which is the practice of automating and managing the lifecycle of ML models. LLM landscapes are currently composed of platforms (e.g., Vertex AI) to manage end-to-end deployment solutions and frameworks (e.g., LangChain) to customize LLMs integration and application development. This paper attempts to understand the key differences between LLMOps and MLOps, highlighting their unique challenges, infrastructure requirements, and methodologies. The paper explores the distinction between traditional ML workflows and those required for LLMs to emphasize security concerns, scalability, and ethical considerations. Fundamental platforms, tools, and emerging trends in LLMOps are evaluated to offer actionable information for practitioners. Finally, the paper presents future potential trends for LLMOps by focusing on its critical role in optimizing LLMs for production use in fields such as healthcare, finance, and cybersecurity. Full article

(This article belongs to the Special Issue Natural Language Processing (NLP) with Applications and Natural Language Understanding (NLU))

► Show Figures

Figure 1

15 pages, 3271 KiB

Open AccessArticle

Evaluation of Academic Stress Employing Network and Time Series Analysis on EEG Data

by Ioannis Konstantinidis, Spyros Avdimiotis and Theodosios Sapounidis

Information 2025, 16(2), 86; https://doi.org/10.3390/info16020086 - 23 Jan 2025

Abstract

Stress management during examinations is an issue perpetually concerning students at all levels of education. This research endeavor focuses not merely on the plethora of factors associated with stress during examinations but rather focuses primarily on specific emotions developed during written and oral [...] Read more.

Stress management during examinations is an issue perpetually concerning students at all levels of education. This research endeavor focuses not merely on the plethora of factors associated with stress during examinations but rather focuses primarily on specific emotions developed during written and oral exams, introducing an original method, based on electroencephalograph (EEG) data. In an attempt to address the role of emotions in stress management, we organized an experiment on a sample of 30 postgraduate students, administering the Trier research protocol. The validated Emotiv Epoc+ device was used, while acquired the data were processed using a combined methodology of network and time series analysis applied in three (3) steps. Initially a descriptive analysis was performed based on the extracted frequencies, while in the second phase a network analysis was conducted to understand the extend of coordination between measured emotions (stress, engagement, interest, relaxation, excitement, and focus), using the tools of centrality. Finally, time series analysis was employed to indicate the level of auto- and cross-correlation between stress and other emotions. Our findings show that (eu)stress plays an imperative role in examination success, as long as it normally fluctuates within middle-range values (on a scale 0–100), and is being restrained by other feelings, such as interest, excitement, and focus Full article

(This article belongs to the Section Biomedical Information and Health)

► Show Figures

Graphical abstract

18 pages, 1031 KiB

Open AccessArticle

An Ensemble Framework for Text Classification

by Eleni Kamateri and Michail Salampasis

Information 2025, 16(2), 85; https://doi.org/10.3390/info16020085 - 23 Jan 2025

Abstract

Ensemble learning can improve predictive performance compared to the performance of any of its constituents alone, while keeping computational demands manageable. However, no reference methodology is available for developing ensemble systems. In this paper, we adapt an ensemble framework for patent classification to [...] Read more.

Ensemble learning can improve predictive performance compared to the performance of any of its constituents alone, while keeping computational demands manageable. However, no reference methodology is available for developing ensemble systems. In this paper, we adapt an ensemble framework for patent classification to assist data scientists in creating flexible ensemble architectures for text classification by selecting a finite set of constituent base models from the many available alternatives. We analyze the axes along which someone can select base models of an ensemble system and propose a methodology for combining them. Moreover, we conduct experiments to compare the effectiveness of ensemble systems against base models and state-of-the-art methods on multiple datasets (three patent classification and two text classification datasets), including long and short texts and single- and/or multi-labeled texts. The results verify the generality of our framework and the effectiveness of ensemble systems, especially ensembles of classifiers trained on different data sections/metadata. Full article

(This article belongs to the Special Issue Machine Learning and Data Mining: Innovations in Big Data Analytics)

► Show Figures

Graphical abstract

15 pages, 293 KiB

Open AccessArticle

Fine-Tuning QurSim on Monolingual and Multilingual Models for Semantic Search

by Tania Afzal, Sadaf Abdul Rauf, Muhammad Ghulam Abbas Malik and Muhammad Imran

Information 2025, 16(2), 84; https://doi.org/10.3390/info16020084 - 23 Jan 2025

Abstract

Transformers have made a significant breakthrough in natural language processing. These models are trained on large datasets and can handle multiple tasks. We compare monolingual and multilingual transformer models for semantic relatedness and verse retrieval. We leveraged data from the original QurSim dataset [...] Read more.

Transformers have made a significant breakthrough in natural language processing. These models are trained on large datasets and can handle multiple tasks. We compare monolingual and multilingual transformer models for semantic relatedness and verse retrieval. We leveraged data from the original QurSim dataset (Arabic) and used authentic multi-author translations in 22 languages to create a multilingual QurSim dataset, which we released for the research community. We evaluated the performance of monolingual and multilingual LLMs for Arabic and our results show that monolingual LLMs give better results for verse classification and matching verse retrieval. We incrementally built monolingual models with Arabic, English, and Urdu and multilingual models with all 22 languages supported by the multilingual paraphrase-MiniLM-L12-v2 model. Our results show improvement in classification accuracy with the incorporation of multilingual QurSim. Full article

(This article belongs to the Special Issue Methods for Integrating Information in Data, Language Models, and Knowledge Graphs for Neurosymbolic Learning and Reasoning)

► Show Figures

Figure 1

25 pages, 2777 KiB

Open AccessArticle

A Modifiable Blockchain Based on the RE-TNG Node Selection Method

by Rongtao Chen, Chao Li, Bingrong Dai and Shaohua Zhang

Information 2025, 16(2), 83; https://doi.org/10.3390/info16020083 - 23 Jan 2025

Abstract

Blockchain technology, characterized by its immutability and decentralization, enables the creation of permanent and tamper-resistant records once data are uploaded, making it widely applicable in scenarios requiring data authenticity and reliability. However, the immutability of on-chain data poses significant security risks, as erroneous [...] Read more.

Blockchain technology, characterized by its immutability and decentralization, enables the creation of permanent and tamper-resistant records once data are uploaded, making it widely applicable in scenarios requiring data authenticity and reliability. However, the immutability of on-chain data poses significant security risks, as erroneous or illegal data become difficult to correct or remove once recorded. Editable blockchain technology offers a potential solution for on-chain data modification. Nevertheless, existing approaches face several challenges, including the impact of malicious nodes on the security and efficiency of data modification, excessive centralization in the management of modification rights and trapdoor keys, and cumulative issues in reputation-based traditional node grouping methods. To address these challenges, this study proposes an RE-TNG (Reputation Evaluation-Twice Node Grouping) node selection method and an editable blockchain scheme based on it. The RE-TNG method employs a two-stage grouping process following reputation-based node ranking. The first grouping stage uses a Fibonacci sequence-based rule to mitigate the issue of cumulative reputation values over time. The second grouping stage selects high-reputation nodes within groups to ensure the trustworthiness of selected nodes. Trapdoor keys are collaboratively generated by the high-reputation node group, achieving decentralized trapdoor management. Modification nodes are randomly chosen from the high-reputation group, ensuring both integrity and decentralization in modification authority. Comparative analyses and experimental evaluations against traditional random node selection and grouping methods demonstrate the feasibility of the proposed scheme, showcasing a superior performance in terms of security and modification efficiency. Full article

► Show Figures

Figure 1

28 pages, 724 KiB

Open AccessArticle

Determinants of ChatGPT Adoption Intention in Higher Education: Expanding on TAM with the Mediating Roles of Trust and Risk

by Stefanos Balaskas, Vassilios Tsiantos, Sevaste Chatzifotiou and Maria Rigou

Information 2025, 16(2), 82; https://doi.org/10.3390/info16020082 - 22 Jan 2025

Abstract

Generative AI, particularly tools like ChatGPT, is reshaping higher education by enhancing academic engagement, streamlining processes, and fostering innovation. This study investigates the determinants of ChatGPT adoption intentions (CGPTAIs) by extending the Technology Acceptance Model (TAM) to include the mediating roles of perceived [...] Read more.

Generative AI, particularly tools like ChatGPT, is reshaping higher education by enhancing academic engagement, streamlining processes, and fostering innovation. This study investigates the determinants of ChatGPT adoption intentions (CGPTAIs) by extending the Technology Acceptance Model (TAM) to include the mediating roles of perceived trust (PT) and perceived risk (PR). Using a quantitative cross-sectional design, the data from 435 participants were analyzed using structural equation modeling (SEM) to explore the relationships among the perceived ease of use (PE), perceived intelligence (PI), perceived usefulness (PUSE), PT, and PR. Τhe findings reveal that the perceived ease of use (PE) and perceived intelligence (PI) significantly drive adoption intentions, while perceived usefulness (PUSE) plays a limited role. PR fully mediates the relationship between PUSE and CGPTAI and partially mediates PE and PI, while PT fully mediates PUSE and partially mediates PE, but not PI. Multi-group analysis highlights demographic differences, such as age and prior AI experience, in adoption pathways. These results challenge traditional TAM assumptions, advancing the model to account for the interplay of usability, intelligence, trust, and risk. Practical insights are provided for fostering ethical and responsible ChatGPT integration, safeguarding academic integrity, and promoting equitable access in higher education. Full article

(This article belongs to the Special Issue Generative AI Technologies: Shaping the Future of Higher Education)

► Show Figures

Graphical abstract

25 pages, 19156 KiB

Open AccessArticle

Data Augmentation in Earth Observation: A Diffusion Model Approach

by Tiago Sousa, Benoît Ries and Nicolas Guelfi

Information 2025, 16(2), 81; https://doi.org/10.3390/info16020081 - 22 Jan 2025

Abstract

High-quality Earth Observation (EO) imagery is essential for accurate analysis and informed decision making across sectors. However, data scarcity caused by atmospheric conditions, seasonal variations, and limited geographical coverage hinders the effective application of Artificial Intelligence (AI) in EO. Traditional data augmentation techniques, [...] Read more.

High-quality Earth Observation (EO) imagery is essential for accurate analysis and informed decision making across sectors. However, data scarcity caused by atmospheric conditions, seasonal variations, and limited geographical coverage hinders the effective application of Artificial Intelligence (AI) in EO. Traditional data augmentation techniques, which rely on basic parameterized image transformations, often fail to introduce sufficient diversity across key semantic axes. These axes include natural changes such as snow and floods, human impacts like urbanization and roads, and disasters such as wildfires and storms, which limits the accuracy of AI models in EO applications. To address this, we propose a four-stage data augmentation approach that integrates diffusion models to enhance semantic diversity. Our method employs meta-prompts for instruction generation, vision–language models for rich captioning, EO-specific diffusion model fine-tuning, and iterative data augmentation. Extensive experiments using four augmentation techniques demonstrate that our approach consistently outperforms established methods, generating semantically diverse EO images and improving AI model performance. Full article

► Show Figures

Figure 1

Journal Description

Information

Latest Articles

Journal Menu

Journal Browser

Highly Accessed Articles

Latest Books

E-Mail Alert

News

Topics

Conferences

Special Issues

Topical Collections

Further Information

Guidelines

MDPI Initiatives

Follow MDPI