Information

19 pages, 1235 KB

Open AccessArticle

A Novel Optimized Ranking Function Method of Spherical Fuzzy Multi-Attribute Decision-Making and Application

by Haiping Ren, Supan Yang, Jiajie Shi and Tonghua Yang

Information 2026, 17(7), 653; https://doi.org/10.3390/info17070653 (registering DOI) - 4 Jul 2026

Compared to intuitive fuzzy sets (IFSs) and Pythagorean fuzzy sets, spherical fuzzy sets (SFSs) more accurately capture the fuzziness and uncertainty inherent in complex problems. SFSs have found extensive applications across numerous fields. To address the common shortcomings of insufficient ranking capabilities and [...] Read more.

Compared to intuitive fuzzy sets (IFSs) and Pythagorean fuzzy sets, spherical fuzzy sets (SFSs) more accurately capture the fuzziness and uncertainty inherent in complex problems. SFSs have found extensive applications across numerous fields. To address the common shortcomings of insufficient ranking capabilities and unknown attribute weights in existing ranking methods under spherical fuzzy (SF) environments. By integrating the principles of technique for order preference by similarity to an ideal solution (TOPSIS) method into the ranking function design, an innovative spherical fuzzy number (SFN) ranking function is developed that combines the characteristics of a ranking function while retaining the advantages of the TOPSIS method. Secondly, for scenarios where attribute weight information is completely unknown, this paper develops a weigh calculation method based on the newly proposed ranking function combined with the criteria importance through the intercriteria correlation (CRITIC) method. Using the Lagrange multiplier method to construct an optimization model, a specific formula for determining attribute weights is derived. For scenarios with partially known attribute weights, an optimization model is constructed where the optimal solution corresponds to the desired attribute weight values. Finally, two case studies validate the feasibility and effectiveness of the new ranking function for SFNs, providing theoretical support for addressing multi-attribute decision-making (DM) problems in complex uncertain environments. Full article

(This article belongs to the Special Issue Explainable and Trustworthy AI Through Fuzzy Logic)

31 pages, 1937 KB

Open AccessArticle

Benchmarking Energy Efficiency of Supervised Machine Learning Models on Multi-Domain Classification Datasets

by Aamir Ali, Rohail Qamar, Raheela Asif and Saman Hina

Information 2026, 17(7), 652; https://doi.org/10.3390/info17070652 (registering DOI) - 4 Jul 2026

Abstract

Machine learning should be judged by how well it predicts, and computational resources are not accounted for in predictive accuracy. Given the growing emphasis on energy consumption and resource efficiency, decision-supporting frameworks should go beyond accuracy. This study presents an energy-based benchmarking approach [...] Read more.

Machine learning should be judged by how well it predicts, and computational resources are not accounted for in predictive accuracy. Given the growing emphasis on energy consumption and resource efficiency, decision-supporting frameworks should go beyond accuracy. This study presents an energy-based benchmarking approach for supervised learning models. Ten classical algorithms were evaluated on three textual and tabular datasets. The energy consumption of preprocessing, training, and inference was monitored with Intel RAPL via pyRAPL along with the runtime, peak memory usage, and predictive performance statistics (accuracy, precision, recall, F1-score, and AUC). Experiments were conducted in a controlled CPU-based environment to ensure comparability. The computational role of this feature is found to be appreciably diverse. Results show that Random Forest achieved the highest overall balance between predictive performance and efficiency (CI = 0.950, PPI = 0.907), while Logistic Regression provided a competitive trade-off (CI = 0.905, EI = 0.998). Gaussian Naïve Bayes was the most energy-efficient model with a mean energy consumption of 127 J, whereas Support Vector Classifier (SVC) incurred the highest computational cost, consuming 45,758 J and requiring 3925 s on average. The Pareto analysis identified Random Forest, Logistic Regression, Passive Aggressive, and Decision Tree as non-dominated solutions. These findings demonstrate that accuracy alone can be misleading for model evaluation and that integrating energy, runtime, and memory metrics enables more sustainable and resource-aware machine learning model selection. The proposed framework provides practical guidance for Green AI, Tiny Machine Learning (TinyML), edge computing, and other resource-constrained deployment environments. Full article

(This article belongs to the Special Issue Innovative Machine Learning Technologies and Applications)

27 pages, 4800 KB

Open AccessArticle

Collaborative Governance of Involutionary Competition in Platform Economy Under Traffic Contestation: A Case Study of China’s Food Delivery Platforms

by Yanhong Ma and Yumeng Zhong

Information 2026, 17(7), 651; https://doi.org/10.3390/info17070651 (registering DOI) - 4 Jul 2026

Abstract

The entry of JD.com into the food delivery sector and the ensuing subsidy competition have resulted in irrational competition, merchant profit squeezes, and food safety risks in China. This study therefore investigates the collaborative governance mechanisms for food delivery platforms under involutionary competition [...] Read more.

The entry of JD.com into the food delivery sector and the ensuing subsidy competition have resulted in irrational competition, merchant profit squeezes, and food safety risks in China. This study therefore investigates the collaborative governance mechanisms for food delivery platforms under involutionary competition driven by traffic contestation. A two-agent evolutionary game model between platforms and merchants is developed, and Q-learning simulations are conducted to capture dynamic learning behaviors. The analysis examines the effects of coupon face value, cost-sharing mechanisms, traffic incentives, and government incentive-penalty policies on the strategic choices of both agents. Key findings reveal that merchants are more sensitive than platforms to traffic incentives and government penalties. Traffic-dependent merchants and traffic-independent merchants exhibit significantly different responses to government interventions. The coupon face value demonstrates a threshold effect, where only a reasonable range encourages compliant behavior among both parties. Based on these results, a collaborative governance framework is proposed. For traffic-dependent merchants, the government should focus on regulating platform behaviors and supervising coupon value controls, while platforms should establish a reward-oriented, penalty-supported incentive mechanism. For traffic-independent merchants, the government should strengthen consumer-reporting penalty mechanisms and strictly control collusion risks between platforms and merchants. Platforms should increase inspection frequency and reinforce penalties to prevent, at the source, the decline in product quality and market disorder induced by involutionary competition. This study provides strategic insights for achieving collaborative governance of involutionary competition in platform economies under intense traffic contestation. Full article

(This article belongs to the Special Issue Decision-Making Process in E-Commerce and Social Networks)

► Show Figures

Graphical abstract

22 pages, 3642 KB

Open AccessArticle

A Deployment-Oriented Case Study of YOLO-Based Model Compression for On-Board Space Debris Detection

by Liam Kerr and Ognjen Arandjelović

Information 2026, 17(7), 650; https://doi.org/10.3390/info17070650 - 3 Jul 2026

Abstract

Space debris presents a growing operational risk to spacecraft, especially in low Earth orbit, where collisions can generate further debris and increase future collision probability. Active debris removal and in-orbit servicing require robust close-range perception, but on-board systems are constrained by power, memory, [...] Read more.

Space debris presents a growing operational risk to spacecraft, especially in low Earth orbit, where collisions can generate further debris and increase future collision probability. Active debris removal and in-orbit servicing require robust close-range perception, but on-board systems are constrained by power, memory, processing capability and the need for reliable real-time operation. This paper investigates convolutional object detection for on-board space debris detection using the SPARK 2022 spacecraft detection dataset. A YOLOv3 detector is fine-tuned and used to evaluate post-training compression through static quantisation and pruning. A lightweight architectural variant, YOLO-DWSC, is also introduced by replacing the YOLOv3-tiny backbone convolutions with depthwise separable convolutions while retaining the detection head. The full-precision YOLOv3 model achieves 0.972 mAP₅₀ and 0.884 mAP_50:95, while 8-bit static quantisation reduces model size from 405 MB to 102 MB with only a small reduction in mAP₅₀, although tighter localisation accuracy is more affected. YOLO-DWSC is much smaller and faster, reaching 256.4 FPS on the tested GPU at 43 MB, but with reduced accuracy. We present this work as a controlled case study rather than an attempt at state-of-the-art SPARK 2022 performance. The original challenge test labels were unavailable, and the experiments therefore use a class-balanced re-split of the labelled data. The results should consequently be interpreted as internally controlled comparisons of compression behaviour, not as leaderboard-comparable benchmark results. Pruning and a two-pass refinement method are also evaluated. The results indicate that simple compression methods can be useful for broad region-of-interest detection, but they also show that claims about on-board deployment require caution. Speed benefits are hardware- and runtime-dependent, and safety-critical proximity operations require evaluation criteria better aligned with full-object containment. Full article

(This article belongs to the Special Issue Computer Vision and Image Processing: Technologies and Applications for Multimedia Systems)

► Show Figures

Figure 1

33 pages, 6713 KB

Open AccessArticle

Examining the Performance of Large Language Models in Health Information Quality Evaluation Tasks: An AI Post-Alignment Perspective

by Shenghui Zhang, Jun Zhang, Peng Li, Min Zuo and Lichao Feng

Information 2026, 17(7), 649; https://doi.org/10.3390/info17070649 - 2 Jul 2026

Abstract

When large language models (LLMs) are implemented in the field of health information quality evaluation, ensuring that their outputs align with human judgment, logic, and strategic preferences has become a focal point of current research on AI alignment. This study proposes the Health [...] Read more.

When large language models (LLMs) are implemented in the field of health information quality evaluation, ensuring that their outputs align with human judgment, logic, and strategic preferences has become a focal point of current research on AI alignment. This study proposes the Health Information Quality Evaluation Framework based on AI Post-Alignment (HIQE-PA) to examine whether large language models remain consistent with expert judgment standards in health information quality evaluation tasks. Using expert evaluation results as the benchmark, the framework assesses the consistency, stability, and cross-source adaptability of model outputs through structured task implementation and multidimensional statistical indicators. In the experimental study, ERNIE-3.5 and ChatGLM2-6B-32K were selected as general-purpose models, and the corresponding ERNIE-3.5 + HIQE-PA model and ChatGLM2-6B-32K + HIQE-PA model were constructed to compare post-alignment performance under different model conditions. Under three post-alignment performance evaluation indicators, including deviation degree, predictability, and fitness, the LLMs’ post-alignment performance was examined through their outputs in the health information quality evaluation task. The results show that the general LLMs achieved a high level of post-alignment performance only in the dimensions of readability and completeness, while the LLMs + HIQE-PA improved post-alignment performance across all dimensions. In particular, the ERNIE-3.5 + HIQE-PA model performed prominently, producing evaluation outputs that were closer to expert consensus and maintaining consistency with the expert benchmark across different text sources. This study demonstrates that post-alignment examinations can provide empirical support for model selection and governance in the health information domain. Full article

(This article belongs to the Special Issue Data Mining and Healthcare Informatics)

► Show Figures

Figure 1

20 pages, 722 KB

Open AccessArticle

SenScanner: An Artificial Intelligence-Based Automatic Password-Related Secret Detection System in Mixed Texts

by Zhuofeng He, Yumeng Guo, Bo Zhang and Wenzhi Cao

Information 2026, 17(7), 648; https://doi.org/10.3390/info17070648 (registering DOI) - 2 Jul 2026

Abstract

The rapid expansion of the Internet has enabled large-scale information sharing while also increasing the risk of sensitive information leakage. Authentication secrets, including passwords and API keys, may be unintentionally exposed in publicly accessible environments such as web pages, network packets, and code-sharing [...] Read more.

The rapid expansion of the Internet has enabled large-scale information sharing while also increasing the risk of sensitive information leakage. Authentication secrets, including passwords and API keys, may be unintentionally exposed in publicly accessible environments such as web pages, network packets, and code-sharing platforms when they are mishandled by developers or operators. Such leakage allows attackers to abuse third-party authentication services and may lead to unauthorized access, fraud, or broader compromise. Therefore, timely and accurate detection of sensitive information in network data is essential for reducing security risks. This paper presents SenScanner, an artificial intelligence-based model for automatically identifying password-related secrets in mixed text. By combining natural language processing and machine learning techniques, SenScanner detects leaked password-related sensitive information across heterogeneous textual contexts. Experimental results on 2000 public data samples show that SenScanner achieves superior precision, recall, and F1-score, demonstrating its effectiveness in reducing false positives and manual review effort. Full article

(This article belongs to the Special Issue AI-Driven Information Analytics for Cybersecurity and Privacy)

► Show Figures

Graphical abstract

26 pages, 2049 KB

Open AccessSystematic Review

Systematic Review of Privacy Preservation in Federated Learning for Secured Healthcare Applications

by Anu Alankamony and Ninisha Nels

Information 2026, 17(7), 647; https://doi.org/10.3390/info17070647 - 2 Jul 2026

Abstract

The quick transition of the healthcare industry to digital during the era of the Internet of Medical Things and Artificial Intelligence has ignited the demand for frameworks for data sharing while retaining safety and patient privacy. Centralized learning models place potentially sensitive patient [...] Read more.

The quick transition of the healthcare industry to digital during the era of the Internet of Medical Things and Artificial Intelligence has ignited the demand for frameworks for data sharing while retaining safety and patient privacy. Centralized learning models place potentially sensitive patient data at risk of leakage, regulatory violation, and cyber-attacks which undermine receptivity and responsible ownership of big medical data. Federated learning is a novel paradigm that allows patients from various healthcare entities to train machine learning models while maintaining the ability to leverage their data without sharing their direct data. This study proposes a systematic literature review of approaches of privacy-preserving federated learning frameworks in healthcare applications. Following PRISMA guidelines, searches were conducted across Web of Science, Scopus, IEEE Xplore, ScienceDirect, PubMed, and ACM Digital Library with predefined query strings, explicit inclusion/exclusion criteria, and quality appraisal procedures. A total of 80 peer-reviewed studies, published from January 2015 to December 2025, were included in this systematic review, which examined cryptographic, architectural and algorithmic methods including differential privacy, homomorphic encryption, and Secure Multi-Party Computation, along with integrations using blockchain to enhance trust and confidence in distributed healthcare systems. The findings indicate a gradual shift towards hybrid privacy-preserving federated learning architectures which combined multiple security mechanisms to improve trust, confidentiality and robustness. Although significant progress has been achieved, the real-world deployment of such systems is heavily affected due to the challenges in communication efficiency, non-IID data distribution, adversarial attacks, and regulatory requirements. This research highlights future research directions for scalable, explainable and interoperable federated architectures that strike an optimal balance of privacy, utility and system performance for next-gen health intelligence. Trial registration: PROSPERO (CRD420261401073). Full article

(This article belongs to the Special Issue Advanced AI and Data-Driven Learning Methods for Healthcare Applications)

► Show Figures

Figure 1

36 pages, 472 KB

Open AccessFeature PaperArticle

A Novel Enterprise AI Classification Framework for Business Transformation: A Structured Literature Review and Integration of AI Types and Autonomy Levels

by Nusi Borovac and Vladimir Stantchev

Information 2026, 17(7), 646; https://doi.org/10.3390/info17070646 - 1 Jul 2026

Abstract

Enterprise investment in artificial intelligence has reached an unprecedented scale across national, regional, and organisational levels of the economy, yet transformation outcomes remain highly variable. Recent global research indicates that the majority of enterprise AI investments produce no measurable profit-and-loss impact, with only [...] Read more.

Enterprise investment in artificial intelligence has reached an unprecedented scale across national, regional, and organisational levels of the economy, yet transformation outcomes remain highly variable. Recent global research indicates that the majority of enterprise AI investments produce no measurable profit-and-loss impact, with only a minority of organisations extracting material enterprise-level value. Without an integrated classification framework that allows enterprises to deploy, govern, and create value from the full range of AI technologies and autonomy levels across enterprise functions, the risks of misallocated investment, fragmented deployment, and unrealised return remain difficult to mitigate. A structured literature review on AI classification reveals that existing frameworks do not integrate AI technology types, operational autonomy levels, and concrete enterprise applications in a single classification structure that can support enterprise AI-driven transformation decisions. To address this gap, this paper proposes the Enterprise AI Classification Framework, a novel classification model that integrates six AI technology types with six autonomy levels in a 6 × 6 matrix of 36 combinations, each corresponding to concrete business applications across enterprise functions. A computational pilot study (n = 20 cases for case-based coding validation and 12 LLM-simulated personas × 30 vignettes for role-specific hypothesis pre-specification) provides preliminary evidence of substantially higher classification coverage than four baseline frameworks and pre-specifies role-divergence hypotheses for a planned empirical validation, with all design choices locked in an OSF pre-registration; reliability point estimates exceed the pre-specified

α \geq 0.70

threshold at the pilot scale, with bootstrapped confidence intervals expected to tighten under the full-corpus run in the follow-up study. An interview-based empirical validation with approximately 250 decision-makers and practitioners across 20 industries is in preparation. Full article

(This article belongs to the Topic Advances in Integrative AI, Machine Learning, and Big Data for Transformative Applications)

► Show Figures

Graphical abstract

20 pages, 853 KB

Open AccessArticle

MGGCL: Motif-Guided Graph Contrastive Learning for Recommendation

by Li Pang, Yuqi Zhang, Nancy Wang, Jian Yu and Deling Huang

Information 2026, 17(7), 645; https://doi.org/10.3390/info17070645 - 1 Jul 2026

Abstract

Graph motifs capture crucial structural patterns within user–item interaction graphs and offer meaningful semantics that are typically underutilised in conventional graph-based collaborative filtering. Existing contrastive learning methods in recommender systems rely primarily on simple perturbations, which limits their ability to leverage deeper motif-based [...] Read more.

Graph motifs capture crucial structural patterns within user–item interaction graphs and offer meaningful semantics that are typically underutilised in conventional graph-based collaborative filtering. Existing contrastive learning methods in recommender systems rely primarily on simple perturbations, which limits their ability to leverage deeper motif-based structural information. To address this gap, we propose a novel motif-guided contrastive learning framework for recommender systems. To exploit complementary structural biases inherent to user–item interactions, our approach explicitly incorporates three distinct motifs into the construction of the contrastive view. By contrasting views that capture the structural differences among different motifs, our model learns meaningful group-level relationships and potentially suppresses noise arising from sparse or isolated interactions. Extensive experiments on four real-world recommendation datasets validate that our motif-based contrastive approach achieves the best overall performance, outperforming state-of-the-art baselines on three of the four benchmarks with statistically significant margins on the more skewed datasets, while remaining competitive on the fourth, which demonstrates notable robustness and improved accuracy. Full article

(This article belongs to the Special Issue 2nd Edition of Modern Recommender Systems: Approaches, Challenges and Applications)

28 pages, 3672 KB

Open AccessArticle

EPCF: An Equivariant Positional Propagation Enhanced Graph Neural Network for Collaborative Filtering

by Xin Sun, Jishen Sun, Li Pang, Guiling Wang, Zhizhong Liu, Xin Liu and Jian Yu

Information 2026, 17(7), 644; https://doi.org/10.3390/info17070644 - 1 Jul 2026

Abstract

Graph neural networks (GNNs) have shown great advantages in collaborative filtering recommender systems due to their capacity to model user–item relationships through information propagation. However, traditional GNN-based recommenders often fail to distinguish nodes with the same local structure, leading to identical representations after [...] Read more.

Graph neural networks (GNNs) have shown great advantages in collaborative filtering recommender systems due to their capacity to model user–item relationships through information propagation. However, traditional GNN-based recommenders often fail to distinguish nodes with the same local structure, leading to identical representations after propagation. Some studies address this issue by introducing positional encoding. However, most existing positional encoding approaches break the permutation and orthogonal symmetries of graph representations and degrade generalization ability. To address this limitation, we propose EPCF (equivariant positional collaborative filtering), a novel GNN model for collaborative filtering that introduces an equivariant propagation mechanism for Laplacian positional features. The proposed mechanism preserves equivariance of positional features under orthogonal transformations while maintaining the permutation equivariance inherent to graphs, which can improve generalization. The equivariant positional features are further leveraged to guide node embedding propagation. Our experiments on real-world datasets show that EPCF achieves better average performance than the evaluated baselines, achieving average improvements of 7.01% in Recall@20 and 1.17% in area under the curve (AUC) over the strongest baselines. Furthermore, integrating EPCF as a plug-in mechanism into five different GNN backbone models achieves improvements of 23.13% in Recall@20 and 2.14% in AUC across five datasets, demonstrating its generalization capability. Full article

(This article belongs to the Special Issue Recent Advances in Graph Neural Networks and Their Applications)

► Show Figures

Figure 1

21 pages, 8002 KB

Open AccessArticle

A Lightweight Framework for Android Malware Detection via SDAE-Based Multi-View Static Feature Fusion

by Man Hua, Yanhang Shi and Yanling Li

Information 2026, 17(7), 643; https://doi.org/10.3390/info17070643 - 1 Jul 2026

Abstract

Android malware detection is increasingly important for mobile and edge security because malicious applications may compromise user privacy, device reliability, and sensitive service transactions. However, single-view static detection methods often provide limited semantic coverage and are sensitive to noisy or obfuscated code, while [...] Read more.

Android malware detection is increasingly important for mobile and edge security because malicious applications may compromise user privacy, device reliability, and sensitive service transactions. However, single-view static detection methods often provide limited semantic coverage and are sensitive to noisy or obfuscated code, while many deep learning models remain too heavy for resource-constrained deployment. To address these challenges, this paper proposes a lightweight Android malware detection framework based on SDAE-guided multi-view static feature fusion. The framework extracts three complementary static views, namely API calls, permission requests, and system components, from AndroidManifest.xml and classes.dex. These views are independently denoised and compressed by stacked denoising autoencoders, then aligned as the R, G, and B channels of a pseudo-RGB representation. A compact MicroNet-SE classifier with squeeze-and-excitation blocks is used to recalibrate the fused semantic channels and perform malware classification. Experiments on the CICMalDroid 2020 and CIC-AndMal2017 datasets show that the proposed framework achieves 99.01% accuracy, 99.15% precision, 98.99% recall, and 99.07% F1-score, with only 99.6 k parameters and a model size of 1.26 MB. After conversion to TensorFlow Lite, the MicroNet-SE classifier achieves average on-device inference latencies ranging from 1.85 ms to 2.16 ms on two real mobile devices. The model also maintains stable performance under synthetic feature perturbations and practical APK-level obfuscation settings. These findings suggest that combining multi-view static semantics with denoising-based representation learning can improve both detection robustness and deployment efficiency. Overall, the results indicate that the proposed framework provides an effective and lightweight static screening component for Android malware detection in resource-constrained mobile and edge environments. Full article

(This article belongs to the Section Information Security and Privacy)

► Show Figures

Figure 1

25 pages, 23965 KB

Open AccessArticle

Design, Deployment, and Field Evaluation of a Low-Cost IoT-Based Monitoring System for Urban Particulate Matter: A Winter–Spring Campaign in Almaty, Kazakhstan

by Daniyar Nurseitov, Kairat Bostanbekov, Galymzhan Abdimanap, Raissa Uskenbayeva, Zhuldyz Kalpeyeva and Aiman Moldagulova

Information 2026, 17(7), 642; https://doi.org/10.3390/info17070642 - 1 Jul 2026

Abstract

Air pollution in Almaty, Kazakhstan, poses a critical public health challenge intensified by the city’s basin topography and seasonal thermal inversions that trap anthropogenic emissions. The sparse stationary network (~5 stations for ~2 million inhabitants) lacks the spatial and temporal resolution needed to [...] Read more.

Air pollution in Almaty, Kazakhstan, poses a critical public health challenge intensified by the city’s basin topography and seasonal thermal inversions that trap anthropogenic emissions. The sparse stationary network (~5 stations for ~2 million inhabitants) lacks the spatial and temporal resolution needed to capture intra-urban variability. We present the design, deployment, and field evaluation of a low-cost distributed Internet of Things (IoT) network of six custom nodes—Winsen ZPHS01B multi-parameter modules with Raspberry Pi Zero 2 W edge units, at an estimated principal-component cost of ~US$100 per node—operated during a winter-spring campaign (February–April 2025) and yielding over 70,000 measurements of PM_2.5, PM₁₀, CO₂, temperature, and relative humidity. The system’s novelty lies in three integrated engineering features: an Active Airflow Stabilization enclosure that decouples sampling from external wind, context-aware adaptive edge filtering that reduces transmitted data volume by ~40%, and a secure Edge-DMZ-Core telemetry pipeline. Node readings were cross-validated against a Qingping Air Monitor Pro with documented traceability to FEM-grade reference analyzers (R² = 0.89–0.95), and city-scale consistency was confirmed against the national monitoring dashboard; the network is therefore characterized as providing internally consistent low-cost observations rather than reference-equivalent concentrations. Daily mean PM_2.5 exceeded the WHO 24 h guideline (15 µg/m³) on 84% of monitored days, with February concentrations (54.4 µg/m³) significantly above March (21.9 µg/m³; p < 0.001). A high PM_2.5/PM₁₀ ratio (~0.96), measured at the physically consistent nodes, together with higher weekend concentrations, points to coal-based residential heating as the most likely dominant source. A coupled WRF-SILAM framework is configured for future model-observation integration. The system offers a reproducible, scalable, and cost-effective template for ambient particulate monitoring in resource-constrained cities with complex terrain. Full article

(This article belongs to the Section Internet of Things (IoT))

► Show Figures

Figure 1

25 pages, 8542 KB

Open AccessArticle

MMTR: Strategy-Guided Multimodal Table Reasoning with Reflective Self-Correction

by Lixin Bai, Yibo Ming and Yanmin Chen

Information 2026, 17(7), 641; https://doi.org/10.3390/info17070641 - 1 Jul 2026

Abstract

Although multimodal large language models (MLLMs) have achieved remarkable progress in visual question answering, they remain limited in tabular tasks that require fine-grained structured information perception and complex logical reasoning. This limitation primarily stems from the high density of structured information inherent in [...] Read more.

Although multimodal large language models (MLLMs) have achieved remarkable progress in visual question answering, they remain limited in tabular tasks that require fine-grained structured information perception and complex logical reasoning. This limitation primarily stems from the high density of structured information inherent in tables and the scarcity of high-quality instruction tuning data. To address these challenges and improve the model’s reasoning accuracy in tables, we propose MMTR, a strategy-guided multimodal table reasoning method with reflective self-correction. Mechanistically, we design a dual-LoRA architecture: the Strategy LoRA is responsible for generating structured reasoning steps, while the Reflection LoRA verifies and self-corrects these initial outputs. Their synergy empowers the model with a closed-loop capability of “reasoning–reflection–correction”. On the data front, we construct StrTab-QA, a large-scale dataset comprising question-answering, negative, and reflection samples, providing diverse supervision signals. During training, we further introduce a progressive “reasoning-to-reflection” fine-tuning strategy to gradually achieve cross-modal alignment and structural adaptation. Furthermore, coupled with an adaptive resizing and padding scheme, our approach effectively preserves table structures and minimizes information distortion during visual encoding. Extensive experiments demonstrate that MMTR consistently outperforms strong baselines across multiple table reasoning benchmarks. Full article

(This article belongs to the Section Artificial Intelligence)

► Show Figures

Graphical abstract

31 pages, 1052 KB

Open AccessArticle

Composite Gramian Angular Field for Time-Series Classification

by Pero Bogunović, Saša Mladenović and Andrina Granić

Information 2026, 17(7), 640; https://doi.org/10.3390/info17070640 - 30 Jun 2026

Abstract

Gramian Angular Field (GAF) encodings transform time series into two-dimensional images suitable for convolutional neural network (CNN) classification. Existing applications typically use either the Gramian Angular Summation Field (GASF) or the Gramian Angular Difference Field (GADF) independently, although these two encodings capture complementary [...] Read more.

Gramian Angular Field (GAF) encodings transform time series into two-dimensional images suitable for convolutional neural network (CNN) classification. Existing applications typically use either the Gramian Angular Summation Field (GASF) or the Gramian Angular Difference Field (GADF) independently, although these two encodings capture complementary pairwise angular relationships. This paper proposes the Composite Gramian Angular Field (CGAF), a single-image time-series representation obtained by a weighted algebraic combination of the summation and difference GAF components. The weights are optimised using coarse grid search followed by Gaussian-process Bayesian refinement, with all candidate evaluation restricted to training-only inner validation partitions. The selected weights are frozen before held-out test evaluation. CGAF produces a single encoded output image (approximately 0.08 MB, compared with approximately 0.16 MB for retaining separate GASF and GADF images) and encodes at

5.9 \pm 0.3

ms per sample. We evaluate CGAF in three domain-specific settings—EEG cognitive engagement, PTB-DB heartbeat classification, and FordA automotive fault detection—and on a selected subset of 20 datasets from the UCR Time Series Classification Archive. The method is compared with GASF, GADF, recurrence plots, spectrogram-based encodings, and non-image time-series baselines including SVM, ResNet-1D, InceptionTime, and ROCKET. On the evaluated datasets, CGAF consistently improves over the individual GASF and GADF encodings. It achieves macro-F1

= 0.867 \pm 0.027

on the EEG pilot study, heartbeat-segment-level macro-F1

= 0.941 \pm 0.018

on PTB-DB, and test accuracy

= 91.2 %

on FordA. Because patient identifiers are unavailable for PTB-DB, that result does not establish patient-level generalisation. On the selected UCR subset, CGAF outperforms both GASF and GADF on all 20 datasets. It achieves the best overall accuracy among all evaluated methods on 14 of 20 datasets, whereas ROCKET achieves the best overall accuracy on the remaining six datasets. The results suggest that algebraic integration of summation-based and difference-based angular dependencies can improve image-based time-series classification without modifying the CNN backbone or adding gradient-trained parameters. The EEG results should be interpreted as pilot evidence, whereas broader generalisation requires evaluation on the full UCR/UEA archive, additional biomedical cohorts, and further backbone architectures. Full article

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

26 pages, 1448 KB

Open AccessArticle

Improving Kazakh Abstractive Summarization with LLMs via Domain Adaptation and Self-Correction

by Talgat Zhabayev, Ualsher Tukeyev, Dina Amirova and Nazym Rakhmanberdi

Information 2026, 17(7), 639; https://doi.org/10.3390/info17070639 - 30 Jun 2026

Abstract

Recent advances in large language models (LLMs) have improved abstractive text summarization, yet their benefits remain limited for low-resource and morphologically rich languages such as Kazakh. This study investigates the adaptation of the instruction-tuned Gemma-3-4B-it (Google LLC, Mountain View, CA, USA) model for [...] Read more.

Recent advances in large language models (LLMs) have improved abstractive text summarization, yet their benefits remain limited for low-resource and morphologically rich languages such as Kazakh. This study investigates the adaptation of the instruction-tuned Gemma-3-4B-it (Google LLC, Mountain View, CA, USA) model for Kazakh news summarization. We propose a two-stage training pipeline combining domain-adaptive pretraining (DAPT) on an unlabeled Kazakh news corpus with parameter-efficient supervised fine-tuning (SFT) using LoRA on summarization datasets. We also examine an inference-time self-correction step in which the model revises its initial summaries to improve consistency and reduce factual errors. Experiments are conducted on Kazakh news data, including a translated version of XSum and corpora collected from BAQ and TengriNews. We compare the baseline, SFT-only, DAPT-only, and DAPT + SFT configurations, with and without self-correction. Evaluation is performed using ROUGE, BERTScore, chrF++, human assessment, and qualitative linguistic analysis. The results show that DAPT combined with SFT achieves the strongest overall performance, while self-correction further improves automatic scores in the DAPT + SFT setting. However, human evaluation indicates that metric gains do not always correspond to higher factual faithfulness or informativeness. These findings highlight the need for domain adaptation and careful evaluation when applying LLMs to Kazakh summarization. Full article

(This article belongs to the Section Artificial Intelligence)

► Show Figures

Figure 1

49 pages, 4337 KB

Open AccessArticle

Synthetic Data Augmentation for Robust Classification of Diabetic vs. Non-Diabetic Blood FTIR Spectra

by Ahmed Fadlelmoula, Kirill N. Boldyrev, Margarida Gonçalves, Helena Torres, Susana O. Catarino, Graça Minas and Vitor Carvalho

Information 2026, 17(7), 638; https://doi.org/10.3390/info17070638 - 29 Jun 2026

Abstract

Early detection of diabetes mellitus (DM) is essential for preventing disease progression and improving clinical outcomes. However, developing robust machine learning (ML) models for diabetes diagnosis is often constrained by limited data availability, privacy regulations, and challenges with data sharing. This study investigates [...] Read more.

Early detection of diabetes mellitus (DM) is essential for preventing disease progression and improving clinical outcomes. However, developing robust machine learning (ML) models for diabetes diagnosis is often constrained by limited data availability, privacy regulations, and challenges with data sharing. This study investigates a privacy-preserving synthetic data augmentation framework for classifying diabetic and non-diabetic blood serum samples using Fourier Transform Infrared (FTIR) spectroscopy. Two deep generative approaches, Autoencoders (AEs) and Generative Adversarial Networks (GANs), were evaluated for their ability to generate realistic synthetic FTIR spectra while preserving the statistical and biochemical characteristics of the original dataset. Synthetic datasets generated by the AE and GAN models were assessed using six ML classifiers: Support Vector Machine (SVM), Random Forest (RF), K-Nearest Neighbors (KNN), Gradient Boosting (GB), Logistic Regression (LoR), and Decision Tree (DT). Model performance was evaluated using accuracy, precision, recall, F1-score, Receiver Operating Characteristic (ROC) curves, and Area Under the Curve (AUC). Results showed that AE-generated spectra retained stronger discriminative characteristics and were more easily distinguished from the original spectra, whereas GAN-generated spectra exhibited lower classifier separability, suggesting closer alignment with the original data distribution and greater realism for privacy-oriented data augmentation. Correlation analysis demonstrated high spectral fidelity for both approaches. Compared with the original spectra, AE-generated spectra achieved r = 0.9990 and R² = 0.9999, whereas GAN-generated spectra achieved r = 0.9982 and R² = 0.9965. The most prominent diabetes related spectral variations were observed in the carbohydrate (1000–1200 cm⁻¹), Amide I (~1650 cm⁻¹), and lipid-associated (3000–3500 cm⁻¹) regions. To explore the transferability of the proposed framework, a preliminary experimental feasibility study was conducted using independently acquired whole blood FTIR spectra. The generated spectra showed strong agreement with the measured whole blood spectra, demonstrating the potential applicability of the framework under alternative sampling conditions. Because the experimental cohort included only one diabetic volunteer, this analysis was intended solely as a proof-of-concept assessment of spectral feasibility and methodological transferability, rather than as a validation of diabetes classification performance. Overall, the findings demonstrate that synthetic data generation can effectively augment limited FTIR datasets while preserving privacy and key spectral characteristics. The proposed framework provides a promising foundation for privacy-aware biomedical data augmentation and future development of robust FTIR diabetes screening systems. The results should be interpreted as methodological evidence of feasibility and synthetic data utility rather than as evidence of clinical diagnostic readiness, as the serum dataset remains modest in size and the independent whole-blood experiment was intentionally exploring. Full article

(This article belongs to the Special Issue Innovative Machine Learning Technologies and Applications)

30 pages, 10866 KB

Open AccessArticle

Automotive Production Systems: A Diophantine Simulation Framework with Genetic Algorithm-Driven Stochastic Data Generation

by Devibala Subburaman, Jerzy Szymanski, Marta Zurek and Mithileysh Sathiyanarayanan

Information 2026, 17(7), 637; https://doi.org/10.3390/info17070637 - 29 Jun 2026

Abstract

Discrete production planning under integer constrained resource framework is a challenging issue which requires simultaneous consideration of output maximization, resource efficiency, and balanced resource. This research focuses on simulation of an integer-driven production planning model for an automotive production system. It combines the [...] Read more.

Discrete production planning under integer constrained resource framework is a challenging issue which requires simultaneous consideration of output maximization, resource efficiency, and balanced resource. This research focuses on simulation of an integer-driven production planning model for an automotive production system. It combines the genetic algorithm-based stochastic data generator with a precise Diophantine feasibility enumeration. The genetic algorithm is used as a constraint-aware stochastic specification generator to generate feasible production parameter sets within certain operational constraints. Its main purpose is to create representative production environments for feasibility analysis and not to optimize production. A normalized multi-objective scoring function is presented to address the imbalance in the scales of economic and operational measures. A total of 64,518 feasible automotive production plans were enumerated under engine, tire, labor and budget constraints using the proposed framework. The Pareto-efficient solutions to the cost–output space that were identified, formed a discrete, piecewise Pareto frontier. The best production plan had a total of 83 units with 99% of the labor and tire resources exploited, whereas the budget and engine capacities were not binding. The optimal strategy implies full saturation of the labor capacity (>99%) due to the binding nature of labor as an objective. In practice, a safety buffer can be imposed through the introduction of an upper-bound utilization policy (e.g., 95%), which moves the optimal solution marginally inwards along the Pareto frontier. The analysis of sensitivity to changes in resources of

\pm 10 %

showed the preservation of the Pareto structure and resilient adaptability in the output, which validated the usefulness of the suggested strategy in discrete manufacturing decision support. Full article

(This article belongs to the Section Information Applications)

► Show Figures

Figure 1

23 pages, 8787 KB

Open AccessArticle

Fine-Grained Attribute Analysis of User Satisfaction for NEVs: An Interpretable Importance–Performance Analysis Based on Online Reviews

by Yajie Li and Na Yang

Information 2026, 17(7), 636; https://doi.org/10.3390/info17070636 - 29 Jun 2026

Abstract

The rapid growth of the new energy vehicle (NEV) industry has intensified competition among manufacturers, making user satisfaction a strategic priority. Importance–performance analysis (IPA) provides a viable framework for user satisfaction analysis by examining attribute importance and performance from user reviews. However, existing [...] Read more.

The rapid growth of the new energy vehicle (NEV) industry has intensified competition among manufacturers, making user satisfaction a strategic priority. Importance–performance analysis (IPA) provides a viable framework for user satisfaction analysis by examining attribute importance and performance from user reviews. However, existing attribute importance assessment methods lack transparency, and IPA outputs remain coarse-grained. This study aims to develop an improved IPA framework that equips manufacturers with a data-driven approach to product improvement, derived from analysing user satisfaction through consumer reviews. To achieve this, a constrained priority linear programming model is proposed to enhance interpretability in importance measures, and the traditional IPA framework is extended by further diagnosing underperforming attributes through the integration of negative review analysis and the chi-square test, thereby deriving attribute-specific improvement directions. Empirical analysis of 8424 reviews from the Autohome platform demonstrates the effectiveness of the proposed method, and a cross-platform analysis of 4000 reviews from PCauto further confirms its robustness. Results identify that endurance and handling are priority attributes requiring resource allocation, followed by intelligence, interior, and comfort as attributes that need to maintain their advantages. Full article

(This article belongs to the Section Information Applications)

► Show Figures

Graphical abstract

39 pages, 9252 KB

Open AccessArticle

Design and Pilot Evaluation of a Traceability-Oriented Enterprise Architecture for Advance Salary Payment Management

by Aliya Turegeldinova, Aray Kassenkhan, Bakytzhan Amralinova, Shynara Sarkambayeva, Shyndauyl Nugumanov and Zhainagul Khamitova

Information 2026, 17(7), 635; https://doi.org/10.3390/info17070635 - 29 Jun 2026

Abstract

Advance salary payment services are increasingly used to improve employee financial flexibility; however, their implementation is often constrained by fragmented workflows, disconnected HR–payroll coordination, inconsistent accounting synchronization, and limited transaction traceability. This study proposes and evaluates a Traceability-Oriented Enterprise Architecture (TOEA) for advance [...] Read more.

Advance salary payment services are increasingly used to improve employee financial flexibility; however, their implementation is often constrained by fragmented workflows, disconnected HR–payroll coordination, inconsistent accounting synchronization, and limited transaction traceability. This study proposes and evaluates a Traceability-Oriented Enterprise Architecture (TOEA) for advance salary payment management in enterprise payroll environments. The proposed architecture integrates employee self-service interaction, workflow automation, payroll synchronization, payment processing, audit-oriented monitoring, and enterprise interoperability within a unified digital platform. The study employs a Design Science Research approach and introduces a layered architectural framework together with a unified enterprise data model supporting end-to-end lifecycle traceability through globally unique identifiers. The architecture was implemented as a functional prototype and evaluated through a pilot deployment in a university environment employing more than 1000 staff members. The evaluation included process-based comparison, expert validation involving 16 participants from HR, payroll accounting, and management, scalability assessment under increasing workload scenarios, and qualitative stakeholder feedback analysis. The results indicate improvements in workflow transparency, auditability, payroll synchronization, and organizational coordination, while reducing process fragmentation and manual coordination activities. Expert assessments demonstrated positive perceptions of workflow transparency, traceability, and operational monitoring. The findings suggest that traceability-oriented architectural principles provide an effective foundation for interoperable and auditable salary advance management systems and offer a scalable basis for future integration of intelligent analytics and decision-support services. Full article

(This article belongs to the Section Information Systems)

► Show Figures

Graphical abstract

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