Search Results (337)

The exponential growth of scientific literature demands scalable methods to evaluate large-language-model outputs beyond surface-level fluency. We present a two-phase framework that separates generation from evaluation: a retrieval-augmented generation system first produces candidate abstracts, which are then embedded into semantic co-occurrence graphs and assessed using seven robustness metrics from complex network theory. Two experiments were conducted. The first varied model, embedding and prompt configurations, achieved results showing clear differences in performance; the best family combined gemma-2b-it, a prompt inspired by chain-of-Thought reasoning, and all-mpnet-base-v2, achieving the highest graph-based robustness. The second experiment refined the temperature setting for this family, identifying $\tau =0.2$ as optimal, which stabilized results (sd $=0.12$) and improved robustness relative to retrieval baselines ($\Delta E_G=+0.08$, $\Delta \rho =+0.55$). While human evaluation was limited to a small set of abstracts, the results revealed a partial convergence between graph-based robustness and expert judgments of coherence and importance. Our approach contrasts with methods like GraphRAG and establishes a reproducible, model-agnostic pathway for the scalable quality control of LLM-generated scientific content. Full article

The power system, as a critical national infrastructure, faces stealthy and persistent intrusions from Advanced Persistent Threat (APT) attacks. These attack chains span multiple stages and components, while heterogeneous data sources lack unified semantics, limiting the interpretability of current detection methods. To address this, we combine the W3C PROV-DM standard with power-specific semantics to map generic provenance data into standardized provenance graphs. On this basis, we propose a graph neural network framework that jointly models temporal dependencies and structural features. The framework constructs unified provenance graphs with snapshot partitioning, applies Functional Time Encoding (FTE) for temporal modeling, and employs a graph attention autoencoder with node masking and edge reconstruction to enhance feature representations. Through pooling, graph-level embeddings are obtained for downstream detection. Experiments on two public datasets show that our method outperforms baselines across multiple metrics and exhibits clear inter-class separability. In the context of scarce power-domain APT data, this study improves model applicability and interpretability, and it provides a practical path for provenance graph-based intelligent detection in critical infrastructure protection. Full article

For clinical decision support to work, medical knowledge needs to be easy to find quickly and accurately. Retrieval-Augmented Generation (RAG) systems use big language models and document retrieval to help with diagnostic reasoning, but they could cause hallucinations and have strict privacy rules in healthcare. We tested twelve different types of RAG, such as dense, sparse, hybrid, graph-based, multimodal, self-reflective, adaptive, and security-focused pipelines, on 250 de-identified patient vignettes. We used Precision@5, Mean Reciprocal Rank, nDCG@10, hallucination rate, and latency to see how well the system worked. The best retrieval accuracy (P@5 ≥ 0.68, nDCG@10 ≥ 0.67) was achieved by a Haystack pipeline (DPR + BM25 + cross-encoder) and hybrid fusion (RRF). Self-reflective RAG, on the other hand, lowered hallucinations to 5.8%. Sparse retrieval gave the fastest response (120 ms), but it was not as accurate. We also suggest a single framework for reducing hallucinations that includes retrieval confidence thresholds, chain-of-thought verification, and outside fact-checking. Our findings emphasize pragmatic protocols for the secure implementation of RAG on premises, incorporating encryption, provenance tagging, and audit trails. Future directions encompass the incorporation of clinician feedback and the expansion of multimodal inputs to genomics and proteomics for precision medicine. Full article

This study introduces an explainable neuro-symbolic and large language model (LLM)-driven framework for intelligent interpretation of corneal topography and precision surgical decision support. In a prospective cohort of 20 eyes, comprehensive IOLMaster 700 reports were analyzed through a four-stage pipeline: (1) automated extraction of key parameters—including corneal curvature, pachymetry, and axial biometry; (2) mapping of these quantitative features onto a curated corneal disease and refractive-surgery knowledge graph; (3) Bayesian probabilistic inference to evaluate early keratoconus and surgical eligibility; and (4) explainable multi-model LLM reporting, employing DeepSeek and GPT-4.0, to generate bilingual physician- and patient-facing narratives. By transforming complex imaging data into transparent reasoning chains, the pipeline delivered case-level outputs within ~95 ± 12 s. When benchmarked against independent evaluations by two senior corneal specialists, the framework achieved 92 ± 4% sensitivity, 94 ± 5% specificity, 93 ± 4% accuracy, and an AUC of 0.95 ± 0.03 for early keratoconus detection, alongside an F1 score of 0.90 ± 0.04 for refractive surgery eligibility. The generated bilingual reports were rated ≥4.8/5 for logical clarity, clinical usefulness, and comprehensibility, with representative cases fully concordant with expert judgment. Comparative benchmarking against baseline CNN and ViT models demonstrated superior diagnostic accuracy (AUC = 0.95 ± 0.03 vs. 0.88 and 0.90, p < 0.05), confirming the added value of the neuro-symbolic reasoning layer. All analyses were executed on a workstation equipped with an NVIDIA RTX 4090 GPU and implemented in Python 3.10/PyTorch 2.2.1 for full reproducibility. By explicitly coupling symbolic medical knowledge with advanced language models and embedding explainable artificial intelligence (XAI) principles throughout data processing, reasoning, and reporting, this framework provides a transparent, rapid, and clinically actionable AI solution. The approach holds significant promise for improving early ectatic disease detection and supporting individualized refractive surgery planning in routine ophthalmic practice. Full article

Industrial Internet of Things (IIoT) systems face severe security and trust challenges, particularly under cross-domain data sharing and federated orchestration. We present QuantumTrust-FedChain, a cyber-resilient federated learning framework integrating quantum variational trust modeling, blockchain-backed provenance, and Byzantine-robust aggregation for secure IIoT collaboration in 6G networks. The architecture includes a Quantum Graph Attention Network (Q-GAT) for modeling device trust evolution using encrypted device logs. This consensus-aware federated optimizer penalizes adversarial gradients using stochastic contract enforcement, and a shard-based blockchain for real-time forensic traceability. Using datasets from SWaT and TON IoT, experiments show 98.3% accuracy in anomaly detection, 35% improvement in defense against model poisoning, and full ledger traceability with under 8.5% blockchain overhead. This framework offers a robust and explainable solution for secure AI deployment in safety-critical IIoT environments. Full article

Traditional network threat detection based on signatures is becoming increasingly inadequate as network threats and attacks continue to grow in their novelty and sophistication. Such advanced network threats are better handled by anomaly detection based on Machine Learning (ML) models. However, conventional anomaly-based network threat detection with traditional ML and Deep Learning (DL) faces fundamental limitations. Graph Neural Networks (GNNs) and Transformers are recent deep learning models with innovative architectures, capable of addressing these challenges. Reinforcement learning (RL) can facilitate adaptive learning strategies for GNN- and Transformer-based Intrusion Detection Systems (IDS). However, no systematic literature review (SLR) has jointly analyzed and synthesized these three powerful modeling algorithms in network threat detection. To address this gap, this SLR analyzed 36 peer-reviewed studies published between 2017 and 2025, collectively identifying 56 distinct network threats via the proposed threat classification framework by systematically mapping them to Enterprise MITRE ATT&CK tactics and their corresponding Cyber Kill Chain stages. The reviewed literature consists of 23 GNN-based studies implementing 19 GNN model types, 9 Transformer-based studies implementing 13 Transformer architectures, and 4 RL-based studies with 5 different RL algorithms, evaluated across 50 distinct datasets, demonstrating their overall effectiveness in network threat detection. Full article

Global warming is a critical issue today, largely due to the widespread use of fossil fuels in everyday life. One promising solution to reduce reliance on conventional energy sources is to promote the use of renewable power. In particular, to encourage the use of renewable energy in industrial sectors which involve development and manufacture of the industrial artifacts, there is continuous demand for tracing energy sources within the production processes. However, given a sophisticated industrial product that involves diverse and extensive components and their suppliers, the traceability analysis across its production is a critical challenge for ensuring the full utilization of renewable energy. To alleviate this issue, this paper presents a functional framework to support tracing the usage of renewable energy by integrating the Large Language Models (LLMs) and logic programming across supply chain networks. Specifically, the proposed framework contains the following components: (1) adopting graph-based models to process and manage the extensive information within supply chain networks; (2) using the Retrieval-Augmented Generation (RAG) techniques to support the LLM for processing the information related to supply chain networks and generating relevant responses with structured representations; and (3) presenting a logic programming-based solution to support the traceability analysis of renewable energy regarding the responses from the LLM. As a case study, we use a public dataset to evaluate the proposed framework by comparing it to the RAG-based LLM and its variant. Compared to baseline methods solely relying on LLMs, the experiments show that the proposed framework achieves significant improvement. Full article

The knowledge graph link prediction task currently faces challenges such as insufficient semantic fusion of structured knowledge and unstructured text, limited representation learning of long-tailed entities, and insufficient interpretability of the reasoning process. Aiming at the above problems, this paper proposes a multi-chain fusion reasoning framework to realize accurate link prediction. First, a dual retrieval mechanism based on semantic similarity metrics and embedded feature matching is designed to construct a high-confidence candidate entity set; second, entity-attribute chains, entity-relationship chains, and historical context chains are established by integrating context information from external knowledge bases to generate a candidate entity set. Finally, a self-consistency scoring method fusing type constraints and semantic space alignment is proposed to realize the joint validation of structural rationality and semantic relevance of candidate entities. Experiments on two public datasets show that the method in this paper fully utilizes the ability of multi-chain reasoning and significantly improves the accuracy of knowledge graph link prediction. Full article

With the rapid advance of digital technologies, the service industry has become a key driver of sustainable economic growth and the restructuring of international trade. Drawing on value-added trade flows for five pivotal service industries—construction, air transportation, postal telecommunications, financial intermediation, and education—over 2013–2021, this study examines the spatial evolution of the global service value chain (GSVC). Using social network analysis combined with a Temporal Exponential Random Graph Model (TERGM), we assess the dynamics of the GSVC’ core–periphery structure and identify heterogeneous determinants shaping their spatial networks. The findings are as follows: (1) Exports across the five industries display an “East rising, West declining” pattern, with markedly heterogeneous magnitudes of change. (2) The construction industry is Europe-centered; air transportation exhibits a U.S.–China bipolar structure; postal telecommunications show the most pronounced “East rising, West declining” shift, forming four poles (United States, United Kingdom, Germany, China); financial intermediation contracts to a five-pole core (China, United States, United Kingdom, Switzerland, Germany); and education becomes increasingly multipolar. (3) The GSVC core–periphery system undergoes substantial reconfiguration, with some peripheral economies moving toward the core; the core expands in air transportation, while postal telecommunications exhibit strong regionalization. (4) Digital technology, foreign direct investment, and manufacturing structure promote network evolution, whereas income similarity may dampen it; the effects of economic freedom and labor-force size on spatial network restructuring differ significantly by industry. These results underscore the complex interplay of structural, institutional, and geographic drivers in reshaping GSVC networks and carry implications for fostering sustainable services trade, enhancing interregional connectivity, narrowing global development gaps, and advancing an inclusive digital transformation. Full article

Power-system safety is fundamental to protecting lives and ensuring reliable grid operation. Yet, hierarchical text classification (HTC) methods struggle with domain-dense accident narratives that require cross-sentence reasoning, often yielding limited fine-grained recognition, inconsistent label paths, and weak evidence traceability. We propose EG-HPT (Evidence-Grounded Hierarchy-Aware Prompt Tuning), which augments hierarchical prompt tuning with Global Pointer-based nested-entity recognition and a sentence–entity heterogeneous graph to aggregate cross-sentence cues; label-aware attention selects Top-k evidence nodes and a weighted InfoNCE objective aligns label and evidence representations, while a hierarchical separation loss and an ancestor-completeness constraint regularize the taxonomy. On a HFACS-based power-accident corpus, EG-HPT consistently outperforms strong baselines in Micro-F1, Macro-F1, and path-constrained Micro-F1 (C-Micro-F1), with ablations confirming the contributions of entity evidence and graph aggregation. These results indicate a deployable, interpretable solution for automated risk factor analysis, enabling auditable evidence chains and supporting multi-granularity accident intelligence in safety-critical operations. Full article

Background: Ongoing challenges such as geopolitical conflicts, trade disruptions, economic sanctions, and political instability have underscored the urgent need for large manufacturing enterprises to improve resilience and reduce dependence on global supply chains. Integrating regional and local Small- and Medium-Sized Enterprises (SMEs) has been proposed as a strategic approach to enhance supply chain localization, yet barriers such as limited visibility, qualification hurdles, and integration difficulties persist. Methods: This study proposes a comprehensive knowledge graph driven framework for representing and discovering SMEs, implemented as a proof-of-concept in the U.S. BioPharma sector. The framework constructs a curated knowledge graph in Neo4j, converts it to Resource Description Framework (RDF) format, and aligns it with the Schema.org vocabulary to enable semantic interoperability and enhance the discoverability of SMEs. Results: The developed knowledge graph, consisting of 488 nodes and 11,520 edges, enabled accurate multi-hop SME discovery with query response times under 10 milliseconds. RDF serialization produced 16,086 triples, validated across platforms to confirm interoperability and semantic consistency. Conclusions: The proposed framework provides a scalable, adaptable, and generalizable solution for SME discovery and supply chain localization, offering a practical pathway to strengthen resilience in diverse manufacturing industries. Full article

Background: Supply chain disruptions from natural hazards, geopolitical tensions, or global events, such as the COVID-19 pandemic, can trigger widespread shortages, with particularly severe consequences in healthcare through drug supply interruptions. Existing methods to assess shortage risks include scoring, simulation, and machine learning, but these approaches face limitations in interpretability, scalability, or computational cost. This study explores the application of probabilistic risk assessment (PRA), a method widely used in high-reliability industries, to evaluate pharmaceutical supply chain risks. Methods: We developed the supply chain probabilistic risk assessment framework and tool, which integrates facility-level failure probabilities and flow data to construct and quantify fault trees and network graphs. Using FDA inspection data from drug manufacturing facilities, the framework generates shortage risk profiles, performs uncertainty analysis, and computes importance measures to rank facilities by risk significance. Results: SUPRA quantified 7567 supply chain models in under eight seconds, producing facility-level importance measures and shortage risk profiles that highlight critical vulnerabilities. The tool demonstrated scalability, interpretability, and efficiency compared with traditional simulation-based methods. Conclusions: PRA offers a systematic, data-driven approach for shortage risk assessment in supply chains. SUPRA enables decision-makers to anticipate vulnerabilities, prioritize mitigation strategies, and strengthen resilience in critical sectors such as healthcare. Full article

In this paper, we propose a formulation of the permutation flow and job shop scheduling problems using special recursive functions and show its equivalence to the existing classical formulation. Equivalence is understood in the sense that both ways of defining the problem describe the same set of feasible schedules for each pair of jobs and machine numbers. In this paper, the apparatus of recursive functions is used to describe and solve three problems: permutation flow shop; permutation flow shop with the addition of the ‘and’ predicate extending the machine chain to an acyclic graph; and permutation job shop. The predicate ‘and’ allows the description of the flow shop with assembly operation tasks. Recursive functions have a common domain and range. To calculate an optimal schedule for each of these three problems, a branch and bound method is considered based on a recursive function that implements a job swapping algorithm. The complexity of the optimization algorithm does not increase compared to the non-recursive description of the PFSP. This article presents some results for the calculation of optimal schedules on several test instances. It is expected that the new method, based on the description of recursive functions and their superposition, will be productive for formulating and solving some extensions of scheduling problems that have practical significance. Full article

The global push for electrification and decarbonization has sharply increased demand for critical raw materials—especially lithium and copper—heightening financial and strategic pressures on firms that lead these supply chains. Yet, the systemic financial risks arising from inter-firm interdependencies in this sector remain largely unexplored. This article presents a novel distance-based network framework to analyze systemic risk among the world’s top 15 lithium and copper producers (2020–2024). Firms are represented through standardized vectors of profitability and risk indicators (liquidity–solvency), from which we construct a two-layer similarity network using Euclidean distances. Graph-theoretic tools—including Minimum Spanning Tree, eigenvector centrality, modularity detection, and contagion simulations—reveal the structural properties and transmission pathways of financial shocks. The results show a robust-yet-fragile topology: while stable under minor perturbations, the network is highly vulnerable to failures of central firms. These findings highlight the utility of distance-based network models in uncovering hidden fragilities in critical commodity sectors, offering actionable insights for macroprudential regulators, investors, and corporate risk managers amid growing geopolitical and financial entanglement. Full article

Stealth attacks targeting industrial control systems (ICS) exploit subtle sequences of malicious actions, making them difficult to detect with conventional methods. The OPC Unified Architecture (OPC UA) protocol—now widely adopted in SCADA/ICS environments—enhances OT–IT integration but simultaneously increases the exposure of critical infrastructures to sophisticated cyberattacks. Traditional detection approaches, which rely on instantaneous traffic features and static models, neglect the sequential dimension that is essential for uncovering such gradual intrusions. To address this limitation, we propose a hybrid sequential anomaly detection pipeline that combines Markov chain modeling to capture temporal dependencies with machine learning algorithms for anomaly detection. The pipeline is further augmented by explainability through SHapley Additive exPlanations (SHAP) and causal inference using the PC algorithm. Experimental evaluation on an OPC UA dataset simulating Man-In-The-Middle (MITM) and denial-of-service (DoS) attacks demonstrates that incorporating a second-order sequential memory significantly improves detection: F1-score increases by +2.27%, precision by +2.33%, and recall by +3.02%. SHAP analysis identifies the most influential features and transitions, while the causal graph highlights deviations from the system’s normal structure under attack, thereby providing interpretable insights into the root causes of anomalies. Full article