Information

29 pages, 1260 KB

Open AccessArticle

Modelling Social Attachment and Mental States from Facebook Activity with Machine Learning

by Stavroula Kridera and Andreas Kanavos

Information 2025, 16(9), 772; https://doi.org/10.3390/info16090772 - 5 Sep 2025

Social networks generate vast amounts of data that can reveal patterns of human behaviour, social attachment, and mental states. This paper explores advanced machine learning techniques to detect and model such patterns, focusing on community structures, influential users, and information diffusion pathways. To [...] Read more.

Social networks generate vast amounts of data that can reveal patterns of human behaviour, social attachment, and mental states. This paper explores advanced machine learning techniques to detect and model such patterns, focusing on community structures, influential users, and information diffusion pathways. To address the scale, noise, and heterogeneity of social data, we leverage recent advances in graph theory, natural language processing, and anomaly detection. Our framework combines clustering for community detection, sentiment analysis for emotional state inference, and centrality metrics for influence estimation, while integrating multimodal data—including textual and visual content—for richer behavioural insights. Experimental results demonstrate that the proposed approach effectively extracts actionable knowledge, supporting mental well-being and strengthening digital social ties. Furthermore, we emphasise the role of privacy-preserving methods, such as federated learning, to ensure ethical analysis. These findings lay the groundwork for responsible and effective applications of machine learning in social network analysis. Full article

(This article belongs to the Special Issue Information Extraction and Language Discourse Processing)

► Show Figures

Figure 1

22 pages, 288 KB

Open AccessReview

Generative AI and the Information Society: Ethical Reflections from Libraries

by Molefi Matsieli and Stephen Mutula

Information 2025, 16(9), 771; https://doi.org/10.3390/info16090771 - 5 Sep 2025

Abstract

The integration of generative artificial intelligence (generative AI) into library systems is transforming the global information society, offering new possibilities for improving information access, management, and dissemination. However, these advancements also raise significant ethical concerns, including algorithmic bias, epistemic injustice, intellectual property conflicts, [...] Read more.

The integration of generative artificial intelligence (generative AI) into library systems is transforming the global information society, offering new possibilities for improving information access, management, and dissemination. However, these advancements also raise significant ethical concerns, including algorithmic bias, epistemic injustice, intellectual property conflicts, data privacy breaches, job displacement, the spread of misinformation, and increasing digital inequality. This review critically examines these challenges through the lens of the World Summit on the Information Society (WSIS) Action Line C10, which emphasizes the ethical dimensions of the information society. It argues that while such concerns are global, they are particularly acute in the Global South, where structural barriers such as skills shortages, weak policy frameworks, and limited infrastructure undermine equitable access to AI benefits. The review calls for a more inclusive, transparent, and ethically responsible approach to AI adoption in libraries. It underscores the essential role of libraries as stewards of ethical information practices and advocates for collaborative strategies to ensure that generative AI serves as a tool for empowerment, rather than a driver of deepening inequality in the information society. Full article

40 pages, 796 KB

Open AccessArticle

Entropy-Based Assessment of AI Adoption Patterns in Micro and Small Enterprises: Insights into Strategic Decision-Making and Ecosystem Development in Emerging Economies

by Gelmar García-Vidal, Alexander Sánchez-Rodríguez, Laritza Guzmán-Vilar, Reyner Pérez-Campdesuñer and Rodobaldo Martínez-Vivar

Information 2025, 16(9), 770; https://doi.org/10.3390/info16090770 - 5 Sep 2025

Abstract

This study examines patterns of artificial intelligence (AI) adoption in Ecuadorian micro and small enterprises (MSEs), with an emphasis on functional diversity across value chain activities. Based on a cross-sectional dataset of 781 enterprises and an entropy-based model, it assesses internal variability in [...] Read more.

This study examines patterns of artificial intelligence (AI) adoption in Ecuadorian micro and small enterprises (MSEs), with an emphasis on functional diversity across value chain activities. Based on a cross-sectional dataset of 781 enterprises and an entropy-based model, it assesses internal variability in AI use and explores its relationship with strategic perception and dynamic capabilities. The findings reveal predominant partial adoption, alongside high functional entropy in sectors such as mining and services, suggesting an ongoing phase of technological experimentation. However, a significant gap emerges between perceived strategic use and actual functional configurations—especially among microenterprises—indicating a misalignment between intent and organizational capacity. Barriers to adoption include limited technical skills, high costs, infrastructure constraints, and cultural resistance, yet over 70% of non-adopters express future adoption intentions. Regional analysis identifies both the Andean Highlands and Coastal regions as “innovative,” although with distinct profiles of digital maturity. While microenterprises focus on accessible tools (e.g., chatbots), small enterprises engage in data analytics and automation. Correlation analyses reveal no significant relationship between functional diversity and strategic value or capability development, underscoring the importance of qualitative organizational factors. While primarily descriptive, the entropy-based approach provides a robust diagnostic baseline that can be complemented by multivariate or qualitative methods to uncover causal mechanisms and strengthen policy implications. The proposed framework offers a replicable and adaptable tool for characterizing AI integration and informing differentiated support policies, with relevance for Ecuador and other emerging economies facing fragmented digital transformation. Full article

(This article belongs to the Special Issue Advancing Information Systems Through Artificial Intelligence: Innovative Approaches and Applications)

► Show Figures

Figure 1

33 pages, 725 KB

Open AccessReview

Mapping Blockchain Applications in FinTech: A Systematic Review of Eleven Key Domains

by Tipon Tanchangya, Tapan Sarker, Junaid Rahman, Md Shafiul Islam, Naimul Islam and Kazi Omar Siddiqi

Information 2025, 16(9), 769; https://doi.org/10.3390/info16090769 - 5 Sep 2025

Abstract

Blockchain technology is now a useful tool that FinTech organizations use to increase transparency, optimize activities, and seize new possibilities. This research explores blockchain applications within the FinTech sector. This study systematically explores blockchain applications within the FinTech sector by 164 peer-reviewed articles, [...] Read more.

Blockchain technology is now a useful tool that FinTech organizations use to increase transparency, optimize activities, and seize new possibilities. This research explores blockchain applications within the FinTech sector. This study systematically explores blockchain applications within the FinTech sector by 164 peer-reviewed articles, utilizing the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) framework. The review identifies eleven applications, such as smart contracts, financial inclusion, crowdfunding, digital identity, trade finance, regulatory compliance, insurance, asset management, investment, banking, and lending. A mixed-method strategy, combining quantitative and qualitative content analysis, was applied to examine the adoption and impact of blockchain across these subdomains. It further discusses current challenges such as regulatory ambiguity, interoperability limitations, and cybersecurity threats. This paper provides a consolidated framework of blockchain’s actual application in FinTech subdomains and identifies the main gaps in the existing literature. These results have practical implications for practitioners, researchers, and policymakers who seek to harness blockchain for achieving financial innovation and inclusive growth. Full article

(This article belongs to the Special Issue Decision Models for Economics and Business Management)

► Show Figures

Figure 1

17 pages, 1749 KB

Open AccessArticle

Secure Communication and Dynamic Formation Control of Intelligent Drone Swarms Using Blockchain Technology

by Huayu Li, Peiyan Li, Jing Liu and Peiying Zhang

Information 2025, 16(9), 768; https://doi.org/10.3390/info16090768 - 4 Sep 2025

Abstract

With the increasing deployment of unmanned aerial vehicle (UAV) swarms in scenarios such as disaster response, environmental monitoring, and military reconnaissance, the need for secure and scalable formation control has become critical. Traditional centralized architectures face challenges such as limited scalability, communication bottlenecks, [...] Read more.

With the increasing deployment of unmanned aerial vehicle (UAV) swarms in scenarios such as disaster response, environmental monitoring, and military reconnaissance, the need for secure and scalable formation control has become critical. Traditional centralized architectures face challenges such as limited scalability, communication bottlenecks, and single points of failure in large-scale swarm coordination. To address these issues, this paper proposes a blockchain-based decentralized formation control framework that integrates smart contracts to manage UAV registration, identity authentication, formation assignment, and positional coordination. The system follows a leader–follower structure, where the leader broadcasts formation tasks via on-chain events, while followers respond in real-time through event-driven mechanisms. A parameterized control model based on dynamic angle and distance adjustments is employed to support various formations, including V-shape, line, and circular configurations. The transformation from relative to geographic positions is achieved using Haversine and Euclidean methods. Experimental validation in a simulated environment demonstrates that the proposed method achieves lower communication latency and better responsiveness compared to polling-based schemes, while offering enhanced scalability and robustness. This work provides a feasible and secure decentralized control solution for future UAV swarm systems. Full article

► Show Figures

Figure 1

16 pages, 1471 KB

Open AccessArticle

Leveraging Explainable AI for LLM Text Attribution: Differentiating Human-Written and Multiple LLM-Generated Text

by Ayat A. Najjar, Huthaifa I. Ashqar, Omar Darwish and Eman Hammad

Information 2025, 16(9), 767; https://doi.org/10.3390/info16090767 - 4 Sep 2025

Abstract

The development of generative AI Large Language Models (LLMs) raised the alarm regarding the identification of content produced by generative AI vs. humans. In one case, issues arise when students heavily rely on such tools in a manner that can affect the development [...] Read more.

The development of generative AI Large Language Models (LLMs) raised the alarm regarding the identification of content produced by generative AI vs. humans. In one case, issues arise when students heavily rely on such tools in a manner that can affect the development of their writing or coding skills. Other issues of plagiarism also apply. This study aims to support efforts to detect and identify textual content generated using LLM tools. We hypothesize that LLM-generated text is detectable by machine learning (ML) and investigate ML models that can recognize and differentiate between texts generated by humans and multiple LLM tools. We used a dataset of student-written text in comparison with LLM-written text. We leveraged several ML and Deep Learning (DL) algorithms, such as Random Forest (RF) and Recurrent Neural Networks (RNNs) and utilized Explainable Artificial Intelligence (XAI) to understand the important features in attribution. Our method is divided into (1) binary classification to differentiate between human-written and AI-generated text and (2) multi-classification to differentiate between human-written text and text generated by five different LLM tools (ChatGPT, LLaMA, Google Bard, Claude, and Perplexity). Results show high accuracy in multi- and binary classification. Our model outperformed GPTZero (78.3%), with an accuracy of 98.5%. Notably, GPTZero was unable to recognize about 4.2% of the observations, but our model was able to recognize the complete test dataset. XAI results showed that understanding feature importance across different classes enables detailed author/source profiles, aiding in attribution and supporting plagiarism detection by highlighting unique stylistic and structural elements, thereby ensuring robust verification of content originality. Full article

(This article belongs to the Special Issue Generative AI Transformations in Industrial and Societal Applications)

► Show Figures

Figure 1

18 pages, 1495 KB

Open AccessArticle

Retrieval-Augmented Generation vs. Baseline LLMs: A Multi-Metric Evaluation for Knowledge-Intensive Content

by Aparna Vinayan Kozhipuram, Samar Shailendra and Rajan Kadel

Information 2025, 16(9), 766; https://doi.org/10.3390/info16090766 - 4 Sep 2025

Abstract

(1) Background: The development of Generative Artificial Intelligence (GenAI) is transforming knowledge-intensive domains such as Education. However, Large Language Models (LLMs), which serve as the foundational components for GenAI tools, are trained on static datasets, often producing misleading, factually incorrect, or outdated responses. [...] Read more.

(1) Background: The development of Generative Artificial Intelligence (GenAI) is transforming knowledge-intensive domains such as Education. However, Large Language Models (LLMs), which serve as the foundational components for GenAI tools, are trained on static datasets, often producing misleading, factually incorrect, or outdated responses. Our study explores the performance gains of Retrieval-Augmented LLMs over baseline LLMs while also identifying the trade-off opportunity between smaller-parameter LLMs augmented with user-specific data to larger parameter LLMs. (2) Methods: We experimented with four different LLMs, each with a different number of parameters, to generate outputs. These outputs were then evaluated across seven lexical and semantic metrics to identify performance trends in Retrieval-Augmented Generation (RAG)-Augmented LLMs and analyze the impact of parameter size on LLM performance. (3) Results and Discussions: We have synthesized 968 different combinations to identify this trend with the help of different LLM sizes/parameters: TinyLlama 1.1B, Mistral 7B, Llama 3.1 8B, and Llama 1 13 B. These studies were grouped into two themes: RAG-Augmented LLM percentage improvements to baseline LLMs and compelling trade-off possibilities of RAG-Augmented smaller-parameter LLMs to larger-parameter LLMs. Our experiments show that RAG-Augmented LLMs demonstrate high lexical and semantic scores relative to baseline LLMs. This offers RAG-Augmented LLMs as a compelling trade-off for reducing the number of parameters in LLMs and lowering overall resource demands. (4) Conclusions: The findings outline that by leveraging RAG-Augmented LLMs, smaller-parameter LLMs can perform better or equivalently to large-parameter LLMs, particularly demonstrating strong lexical improvements. They reduce the risks of hallucination and keep the output more contextualized, making them a better choice for knowledge-intensive content in academic and research sectors. Full article

► Show Figures

Figure 1

26 pages, 1256 KB

Open AccessSystematic Review

Toward Decentralized Intelligence: A Systematic Literature Review of Blockchain-Enabled AI Systems

by Mohamad Sheikho Al Jasem, Trevor De Clark and Ajay Kumar Shrestha

Information 2025, 16(9), 765; https://doi.org/10.3390/info16090765 - 3 Sep 2025

Abstract

The convergence of decentralized artificial intelligence (DAI), blockchain technology, and smart contracts is reshaping the design and governance of intelligent systems. As these technologies rapidly evolve, addressing privacy within their architecture, usage models, and associated risks has become increasingly critical. This systematic literature [...] Read more.

The convergence of decentralized artificial intelligence (DAI), blockchain technology, and smart contracts is reshaping the design and governance of intelligent systems. As these technologies rapidly evolve, addressing privacy within their architecture, usage models, and associated risks has become increasingly critical. This systematic literature review examines architectural patterns, governance frameworks, real-world applications, and persistent challenges in DAI systems. It identifies prevailing designs such as federated learning integrated with consensus protocols, smart contract-based incentive mechanisms, and decentralized verification methods. Drawing from a diverse body of recent literature, the review highlights implementations across sectors, including healthcare, finance, IoT, autonomous systems, and intelligent infrastructure, each demonstrating significant contributions to privacy, security, and collaborative innovation. Despite these advancements, DAI systems face ongoing obstacles such as scalability limitations, privacy trade-offs, and difficulties with regulatory compliance. The review emphasizes the need for integrative governance approaches that balance transparency, accountability, incentive alignment, and ethical oversight. These elements are proposed as co-evolving pillars essential to establishing trustworthiness in decentralized AI ecosystems. This work offers a comprehensive review for understanding the current landscape and guiding the development of responsible and effective DAI systems in the Web3 era. Full article

(This article belongs to the Special Issue Blockchain, Technology and Its Application)

► Show Figures

Figure 1

22 pages, 298 KB

Open AccessArticle

AI Integration in Organisational Workflows: A Case Study on Job Reconfiguration, Efficiency, and Workforce Adaptation

by Pedro Oliveira, João M. S. Carvalho and Sílvia Faria

Information 2025, 16(9), 764; https://doi.org/10.3390/info16090764 - 3 Sep 2025

Abstract

This study investigates how the integration of artificial intelligence (AI) transforms job practices within a leading European infrastructure company. Grounded in the Feeling Economy framework, the research explores the shift in task composition following AI implementation, focusing on the emergence of new roles, [...] Read more.

This study investigates how the integration of artificial intelligence (AI) transforms job practices within a leading European infrastructure company. Grounded in the Feeling Economy framework, the research explores the shift in task composition following AI implementation, focusing on the emergence of new roles, required competencies, and the ongoing reconfiguration of work. Using a qualitative, single-case study methodology, data were collected through semi-structured interviews with ten employees and company documentation. Thematic analysis revealed five key dimensions: the reconfiguration of job tasks, the improvement of efficiency and quality, psychological and adaptation challenges, the need for AI-related competencies, and concerns about dehumanisation. Findings show that AI systems increasingly assume repetitive and analytical tasks, enabling workers to focus on strategic, empathetic, and creative responsibilities. However, psychological resistance, fears of job displacement, and a perceived erosion of human interaction present implementation barriers. The study provides theoretical contributions by empirically extending the Feeling Economy and task modularisation frameworks. It also offers managerial insights into workforce adaptation, training needs, and the importance of ethical and emotionally intelligent AI integration. Additionally, this study highlights that the Feeling Economy must address AI’s epistemic risks, emphasising fairness, transparency, and participatory governance as essential for trustworthy, emotionally intelligent, and sustainable AI systems. Full article

► Show Figures

Graphical abstract

16 pages, 599 KB

Open AccessArticle

Exploring Predictive Insights on Student Success Using Explainable Machine Learning: A Synthetic Data Study

by Beatriz Santana-Perera, Carmen García-Barceló, Mauricio González Arcas and David Gil

Information 2025, 16(9), 763; https://doi.org/10.3390/info16090763 - 3 Sep 2025

Abstract

Student success is a multifaceted outcome influenced by academic, behavioral, contextual, and socio-environmental factors. With the growing availability of educational data, machine learning (ML) offers promising tools to model complex, nonlinear relationships that go beyond traditional statistical methods. However, the lack of interpretability [...] Read more.

Student success is a multifaceted outcome influenced by academic, behavioral, contextual, and socio-environmental factors. With the growing availability of educational data, machine learning (ML) offers promising tools to model complex, nonlinear relationships that go beyond traditional statistical methods. However, the lack of interpretability in many ML models remains a major obstacle for practical adoption in educational contexts. In this study, we apply explainable artificial intelligence (XAI) techniques—specifically SHAP (SHapley Additive exPlanations)—to analyze a synthetic dataset simulating diverse student profiles. Using LightGBM, we identify variables such as hours studied, attendance, and parental involvement as influential in predicting exam performance. While the results are not generalizable due to the artificial nature of the data, this study reframes its purpose as a methodological exploration rather than a claim of real-world actionable insights. Our findings demonstrate how interpretable ML can be used to build transparent analytic pipelines in education, setting the stage for future research using empirical datasets and real student data. Full article

(This article belongs to the Special Issue International Database Engineered Applications Symposium, 2nd Edition)

► Show Figures

Figure 1

28 pages, 730 KB

Open AccessSystematic Review

Technologies for Reflective Assessment in Knowledge Building Communities: A Systematic Review

by Paula Rodríguez-Chirino, Calixto Gutiérrez-Braojos, Mario Martínez-Gámez and Carlos Rodríguez-Domínguez

Information 2025, 16(9), 762; https://doi.org/10.3390/info16090762 - 3 Sep 2025

Abstract

Reflective assessment is central to knowledge building, as it enables learners to evaluate and improve their collective ideas. In recent years, a wide range of analytic technologies have been developed to support this process, yet there is still a limited understanding of how [...] Read more.

Reflective assessment is central to knowledge building, as it enables learners to evaluate and improve their collective ideas. In recent years, a wide range of analytic technologies have been developed to support this process, yet there is still a limited understanding of how these tools are designed, implemented, and connected to knowledge building principles. This study addresses this gap through a systematic review of the literature focused on analytic technologies that foster reflective assessment in knowledge building environments. Following a rigorous PRISMA methodology, a search in Web of Science was conducted. Studies that did not meet the inclusion criFteria were excluded, resulting in a final selection of 31 empirical studies. The analysis shows that most analytic tools (e.g., KBDeX, KCA, KBAT) are applied to data generated in Knowledge Forum, supporting students in visualizing, analyzing, and reflecting on the collective knowledge advancement process. The review highlights a growing diversity of tools designed to enhance processes such as idea improvement and epistemic agency. By mapping these contributions, the study provides a clearer understanding of how analytic technologies can be used to strengthen collaborative practices and reflective assessment in knowledge building contexts. Full article

► Show Figures

Graphical abstract

23 pages, 1259 KB

Open AccessArticle

Maieutic, Natural, and Artificial Forms in Automatic Control Case Study

by Luigi Fortuna and Adriano Scibilia

Information 2025, 16(9), 761; https://doi.org/10.3390/info16090761 - 2 Sep 2025

Abstract

Maieutics is a remarkable method for discovering new insights through deep dialogue. Defined as “relating to or resembling the Socratic method of eliciting new ideas from another”, the term originates from the Greek word for “midwifery”—as noted in the Merriam-Webster Dictionary. Recently, maieutics [...] Read more.

Maieutics is a remarkable method for discovering new insights through deep dialogue. Defined as “relating to or resembling the Socratic method of eliciting new ideas from another”, the term originates from the Greek word for “midwifery”—as noted in the Merriam-Webster Dictionary. Recently, maieutics has gained renewed relevance in advanced discussions about artificial intelligence, the nature of the mind, and scientific inquiry. This contribution presents a real and extended dialogue, illustrating the power of the maieutic method in addressing key developments in the field of Automatic Control. Over the past 40 years, the authors have followed a unique intellectual path shaped by this method. Inspired by recent research, they have also applied maieutics in interaction with AI systems—particularly ChatGPT. This experiment aimed to replicate, in a condensed timeframe, the long intellectual journey taken over decades. The preliminary results suggest that although AI systems can retrieve historical information, they struggle to capture the deeper, guiding principles of this journey. The authors also identify a significant concern: while the maieutic approach with ChatGPT can serve as a valuable educational tool, it must be complemented by a strong knowledge of dynamical systems leading to innovative paradigms of learning. Full article

(This article belongs to the Special Issue Learning and Knowledge: Theoretical Issues and Applications)

► Show Figures

Figure 1

18 pages, 568 KB

Open AccessArticle

Beyond Cross-Entropy: Discounted Least Information Theory of Entropy (DLITE) Loss and the Impact of Loss Functions on AI-Driven Named Entity Recognition

by Sonia Pascua, Michael Pan and Weimao Ke

Information 2025, 16(9), 760; https://doi.org/10.3390/info16090760 - 2 Sep 2025

Abstract

Loss functions play a significant role in shaping model behavior in machine learning, yet their design implications remain underexplored in natural language processing tasks such as Named Entity Recognition (NER). This study investigates the performance and optimization behavior of five loss functions—L1, L2, [...] Read more.

Loss functions play a significant role in shaping model behavior in machine learning, yet their design implications remain underexplored in natural language processing tasks such as Named Entity Recognition (NER). This study investigates the performance and optimization behavior of five loss functions—L1, L2, Cross-Entropy (CE), KL Divergence (KL), and the proposed DLITE (Discounted Least Information Theory of Entropy) Loss—within transformer-based NER models. DLITE introduces a bounded, entropy-discounting approach to penalization, prioritizing recall and training stability, especially under noisy or imbalanced data conditions. We conducted empirical evaluations across three benchmark NER datasets: Basic NER, CoNLL-2003, and the Broad Twitter Corpus. While CE and KL achieved the highest weighted F1-scores in clean datasets, DLITE Loss demonstrated distinct advantages in macro recall, precision–recall balance, and convergence stability—particularly in noisy environments. Our findings suggest that the choice of loss function should align with application-specific priorities, such as minimizing false negatives or managing uncertainty. DLITE adds a new dimension to model design by enabling more measured predictions, making it a valuable alternative in high-stakes or real-world NLP deployments. Full article

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

► Show Figures

Figure 1

20 pages, 2241 KB

Open AccessArticle

HarmonyTok: Comparing Methods for Harmony Tokenization for Machine Learning

by Maximos Kaliakatsos-Papakostas, Dimos Makris, Konstantinos Soiledis, Konstantinos-Theodoros Tsamis, Vassilis Katsouros and Emilios Cambouropoulos

Information 2025, 16(9), 759; https://doi.org/10.3390/info16090759 - 1 Sep 2025

Abstract

This paper explores different approaches to harmony tokenization in symbolic music for transformer-based models, focusing on two tasks: masked language modeling (MLM) and melodic harmonization generation. Four tokenization strategies are compared, each varying in how chord information is encoded: (1) as full chord [...] Read more.

This paper explores different approaches to harmony tokenization in symbolic music for transformer-based models, focusing on two tasks: masked language modeling (MLM) and melodic harmonization generation. Four tokenization strategies are compared, each varying in how chord information is encoded: (1) as full chord symbols, (2) separated into root and quality, (3) as sets of pitch classes, and (4) as sets of pitch classes where one is designated as a root. A dataset of over 17,000 lead sheet charts is used to train and evaluate RoBERTa for MLM and GPT-2/BART for harmonization. The results show that chord spelling methods—those breaking chords into pitch-class tokens—achieve higher accuracy and lower perplexity, indicating more confident predictions. These methods also produce fewer token-level errors. In harmonization tasks, chunkier tokenizations (with more information per token) generate chords more similar to the original data, while spelling-based methods better preserve structural aspects such as harmonic rhythm and melody–harmony alignment. Audio evaluations reveal that spelling-based models tend toward more generic pop-like harmonizations, while chunkier tokenizations more faithfully reflect the dataset’s style. Overall, while no single tokenization method dominates across all tasks, different strategies may be preferable for specific applications, such as classification or generative style transfer. Full article

(This article belongs to the Special Issue Machine Learning and Artificial Intelligence with Applications)

► Show Figures

Figure 1

32 pages, 2361 KB

Open AccessArticle

Exploring the Use and Misuse of Large Language Models

by Hezekiah Paul D. Valdez, Faranak Abri, Jade Webb and Thomas H. Austin

Information 2025, 16(9), 758; https://doi.org/10.3390/info16090758 - 1 Sep 2025

Abstract

Language modeling has evolved from simple rule-based systems into complex assistants capable of tackling a multitude of tasks. State-of-the-art large language models (LLMs) are capable of scoring highly on proficiency benchmarks, and as a result have been deployed across industries to increase productivity [...] Read more.

Language modeling has evolved from simple rule-based systems into complex assistants capable of tackling a multitude of tasks. State-of-the-art large language models (LLMs) are capable of scoring highly on proficiency benchmarks, and as a result have been deployed across industries to increase productivity and convenience. However, the prolific nature of such tools has provided threat actors with the ability to leverage them for attack development. Our paper describes the current state of LLMs, their availability, and their role in benevolent and malicious applications. In addition, we propose how an LLM can be combined with text-to-speech (TTS) voice cloning to create a framework capable of carrying out social engineering attacks. Our case study analyzes the realism of two different open-source TTS models, Tortoise TTS and Coqui XTTS-v2, by calculating similarity scores between generated and real audio samples from four participants. Our results demonstrate that Tortoise is able to generate realistic voice clone audios for native English speaking males, which indicates that easily accessible resources can be leveraged to create deceptive social engineering attacks. As such tools become more advanced, defenses such as awareness, detection, and red teaming may not be able to keep up with dangerously equipped adversaries. Full article

► Show Figures

Figure 1

28 pages, 6595 KB

Open AccessArticle

Identifying Individual Information Processing Styles During Advertisement Viewing Through EEG-Driven Classifiers

by Antiopi Panteli, Eirini Kalaitzi and Christos A. Fidas

Information 2025, 16(9), 757; https://doi.org/10.3390/info16090757 - 1 Sep 2025

Abstract

Neuromarketing studies the brain function as a response to marketing stimuli. A large amount of neuromarketing research uses data from electroencephalography (EEG) recordings as a response of individuals’ brains to marketing stimuli, aiming to identify the factors that influence consumer behaviour that they [...] Read more.

Neuromarketing studies the brain function as a response to marketing stimuli. A large amount of neuromarketing research uses data from electroencephalography (EEG) recordings as a response of individuals’ brains to marketing stimuli, aiming to identify the factors that influence consumer behaviour that they cannot articulate or are reluctant to reveal. Evidence suggests that individuals’ processing styles affect their reaction to marketing stimuli. In this study, we propose and evaluate a predictive model that classifies consumers as verbalizers or visualizers based on EEG signals recorded during exposure to verbal, visual, and mixed advertisements. Participants (N = 22) were categorized into verbalizers and visualizers using the Style of Processing (SOP) scale and underwent EEG recording while viewing ads. The EEG signals were preprocessed and the five EEG frequency bands were extracted. We employed three classification models for every set of ads: SVM, Decision Tree, and kNN. While all three classifiers performed around the same, with accuracy between 86 and 93%, during cross-validation SVM proved to be the more effective model, with kNN and Decision Tree showing sensitivity to data imbalances. Additionally, we conducted independent t-tests to look for statistically significant differences between the two classes. The t-tests implicated the Theta frequency band. Therefore, these findings highlight the potential of leveraging EEG-based technology to effectively predict a consumer’s processing style for advertisements and offers practical applications in fields such as interactive content designs and user-experience personalization. Full article

(This article belongs to the Special Issue Human–Computer Interaction in Marketing: Emerging Interfaces, Cognitive Engagement, and Child–Computer Interaction)

► Show Figures

Figure 1

15 pages, 1443 KB

Open AccessArticle

Education Strategy for the Net Generation

by Andrej Flogie, Boris Aberšek and Igor Pesek

Information 2025, 16(9), 756; https://doi.org/10.3390/info16090756 - 1 Sep 2025

Abstract

This paper addresses the urgent need to redefine education strategies for the Net Generation in the context of rapid technological and societal changes. First, the educational challenge is placed within a broader philosophical and cultural framework, focusing on the fluid and evolving nature [...] Read more.

This paper addresses the urgent need to redefine education strategies for the Net Generation in the context of rapid technological and societal changes. First, the educational challenge is placed within a broader philosophical and cultural framework, focusing on the fluid and evolving nature of knowledge and human experience. Building on the paradigm shift from Web 2.0 to Web 4.0 and the emergence of Education 5.0, this paper investigates the pedagogical implications of these developments. Through conceptual analysis supported by contemporary educational theory, this paper proposes a model of education that integrates personalized learning, real-time feedback, and collaborative, interdisciplinary environments. A special focus is placed on the role of educators as mentors, rather than mere transmitters of information, and on the ethical, social, and emotional dimensions of digital learning. This article highlights the importance of adjusting educational practices to real-life contexts and future challenges of young learners while ensuring that the humanistic essence of education is not lost. Full article

(This article belongs to the Special Issue ICT-Based Modelling and Simulation for Education)

► Show Figures

Figure 1

20 pages, 10980 KB

Open AccessArticle

DBN: A Dual-Branch Network for Detecting Multiple Categories of Mental Disorders

by Longhao Zhang, Hongzhen Cui and Yunfeng Peng

Information 2025, 16(9), 755; https://doi.org/10.3390/info16090755 - 31 Aug 2025

Abstract

Mental disorders (MDs) constitute significant risk factors for self-harm and suicide. The incidence of MDs has been increasing annually, primarily due to inadequate diagnosis and intervention. Early identification and timely intervention can effectively slow the progression of MDs and enhance the quality of [...] Read more.

Mental disorders (MDs) constitute significant risk factors for self-harm and suicide. The incidence of MDs has been increasing annually, primarily due to inadequate diagnosis and intervention. Early identification and timely intervention can effectively slow the progression of MDs and enhance the quality of life. However, the high cost and complexity of in-hospital screening exacerbate the psychological burden on patients. Moreover, existing studies primarily focus on the identification of individual subcategories and lack attention to model explainability. These approaches fail to adequately address the complexity of clinical demands. Early screening of MDs using EEG signals and deep learning techniques has demonstrated simplicity and effectiveness. To this end, we constructed a Dual-Branch Network (DBN) leveraging resting-state Quantitative Electroencephalogram (QEEG) features. The DBN is designed to enable the detection of multiple categories of MDs. Firstly, a dual-branch feature extraction strategy was designed to capture multi-dimensional latent features. Further, we propose a Multi-Head Attention Mechanism (MHAM) that integrates dynamic routing. This architecture assigns greater weights to key elements and enhances information transmission efficiency. Finally, the diagnosis is derived from a fully connected layer. In addition, we incorporate SHAP analysis to facilitate feature attribution. This technique elucidates the contribution of significant features to MD detection and improves the transparency of model predictions. Experimental results demonstrate the effectiveness of DBN in detecting various MD categories. The performance of DBN surpasses that of traditional machine learning models. Ablation studies further validate the architectural soundness of DBN. The DBN effectively reduces screening complexity and demonstrates significant potential for clinical applications. Full article

► Show Figures

Figure 1

16 pages, 951 KB

Open AccessArticle

Deep LSTM Surrogates for MEMD: A Noise-Assisted Approach to EEG Intrinsic Mode Function Extraction

by Pablo Andres Muñoz-Gutierrez, Diego Fernando Ramirez-Jimenez and Eduardo Giraldo

Information 2025, 16(9), 754; https://doi.org/10.3390/info16090754 - 31 Aug 2025

Abstract

In this paper, we propose a deep learning-based surrogate model for Multivariate Empirical Mode Decomposition (MEMD) using Long Short-Term Memory (LSTM) networks, aimed at efficiently extracting Intrinsic Mode Functions (IMFs) from electroencephalographic (EEG) signals. Unlike traditional data-driven methods, our approach leverages temporal sequence [...] Read more.

In this paper, we propose a deep learning-based surrogate model for Multivariate Empirical Mode Decomposition (MEMD) using Long Short-Term Memory (LSTM) networks, aimed at efficiently extracting Intrinsic Mode Functions (IMFs) from electroencephalographic (EEG) signals. Unlike traditional data-driven methods, our approach leverages temporal sequence modeling to learn the decomposition process in an end-to-end fashion. We further enhance the decomposition targets by employing Noise-Assisted MEMD (NA-MEMD), which stabilizes mode separation and mitigates mode mixing effects, leading to better supervised learning signals. Extensive experiments on synthetic and real EEG data demonstrate the superior performance of the proposed LSTM surrogate over conventional feedforward neural networks and standard MEMD-based targets. Specifically, the LSTM trained on NA-MEMD outputs achieved the lowest mean squared error (MSE) and the highest signal-to-noise ratio (SNR), significantly outperforming the feedforward baseline, even when compared using the Power Spectral Density (PSD). These results confirm the effectiveness of combining LSTM architectures with noise-assisted decomposition strategies to approximate nonlinear signal analysis tasks such as MEMD. The proposed surrogate model offers a fast and accurate alternative to classical empirical methods, enabling real-time and scalable EEG analysis. Full article

(This article belongs to the Special Issue Signal Processing and Machine Learning, 2nd Edition)

► 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