Search Results (1,113)

Large language models (LLMs) often tend to hallucinate, especially in domain-specific tasks and tasks that require reasoning. Previously, we introduced SubGraph Retrieval Augmented Generation (SG-RAG) as a novel Graph RAG method for multi-hop question answering. SG-RAG leverages Cypher queries to search a given knowledge graph and retrieve the subgraph necessary to answer the question. The results from our previous work showed the higher performance of our method compared to the traditional Retrieval Augmented Generation (RAG). In this work, we further enhanced SG-RAG by proposing an additional step called Merging and Ordering Triplets (MOT). The new MOT step seeks to decrease the redundancy in the retrieved triplets by applying hierarchical merging to the retrieved subgraphs. Moreover, it provides an ordering among the triplets using the Breadth-First Search (BFS) traversal algorithm. We conducted experiments on the MetaQA benchmark, which was proposed for multi-hop question-answering in the movies domain. Our experiments showed that SG-RAG MOT provided more accurate answers than Chain-of-Thought and Graph Chain-of-Thought. We also found that merging (up to a certain point) highly overlapping subgraphs and defining an order among the triplets helped the LLM to generate more precise answers. Full article

(1) Background and objectives: Large language models (LLMs) such as GPT, Mistral, and LLaMA exhibit strong capabilities in text generation, yet assessing the quality of their reasoning—particularly in open-ended and argumentative contexts—remains a persistent challenge. This study introduces Dialectical Agent, an internally developed modular framework designed to evaluate reasoning through a structured three-stage process: opinion, counterargument, and synthesis. The framework enables transparent and comparative analysis of how different LLMs handle dialectical reasoning. (2) Methods: Each stage is executed by a single model, and final syntheses are scored via two independent LLM evaluators (LLaMA 3.1 and GPT-4o) based on a rubric with four dimensions: clarity, coherence, originality, and dialecticality. In parallel, a rule-based semantic analyzer detects rhetorical anomalies and ethical values. All outputs and metadata are stored in a Neo4j graph database for structured exploration. (3) Results: The system was applied to four open-weight models (Gemma 7B, Mistral 7B, Dolphin-Mistral, Zephyr 7B) across ten open-ended prompts on ethical, political, and technological topics. The results show consistent stylistic and semantic variation across models, with moderate inter-rater agreement. Semantic diagnostics revealed differences in value expression and rhetorical flaws not captured by rubric scores. (4) Originality: The framework is, to our knowledge, the first to integrate multi-stage reasoning, rubric-based and semantic evaluation, and graph-based storage into a single system. It enables replicable, interpretable, and multidimensional assessment of generative reasoning—supporting researchers, developers, and educators working with LLMs in high-stakes contexts. Full article

Differential equations need boundary conditions (BCs) for their solution. It is widely acknowledged that differential equations and BCs are representative of independent physical processes, and no correlations between them are required. Two recent studies by Hilhorst, Chung et al. argue instead that, in the specific case of diffusion equations (DEs) in bounded systems, BCs are uniquely constrained by the form of transport coefficients. In this paper, we revisit how DEs emerge as fluid limits out of a picture of stochastic transport. We point out their limits of validity and argue that, in most physical systems, BCs and DEs are actually uncorrelated by virtue of the failure of diffusive approximation near the system’s boundaries. When, instead, the diffusive approximation holds everywhere, we show that the correct chain of reasoning goes in the direction opposite to that conjectured by Hilhorst and Chung: it is the choice of the BCs that determines the form of the DE in the surroundings of the boundary. Full article

Per- and polyfluoroalkyl substances (PFAS) have been found worldwide in water, soil, plants, and animals, including humans. A primary route of exposure for humans and animals to PFAS is through the diet and drinking water. Perfluorooctane sulfonate (PFOS), a long-chain PFAS with a relatively long half-life, has been associated with adverse health effects in humans and laboratory animals. There are few toxicokinetic studies on PFOS in domestic livestock raised for human food consumption, which are critical for assessing human food safety. This work aimed to develop a simple daily accumulation model (DAM) for predicting PFOS residues in edible beef cattle muscle. A one-compartment toxicokinetic model in a spreadsheet format was developed using simple calculations to account for daily PFAS into and out of the animal. The DAM was used to simulate two case studies to predict resultant PFOS residues in edible beef cattle tissues. The results demonstrated that the model can reasonably predict PFOS concentrations in beef cattle muscle in a real-world scenario. The DAM was then used to simulate dietary PFOS exposure in beef cattle throughout a typical lifespan in order to derive a generic bioaccumulation factor. The DAM is expected to work well for other PFAS in beef cattle, PFAS in other livestock species raised for meat, and other chemical contaminants with relatively long half-lives. Full article

In recent years, geographic knowledge graphs (GeoKGs) have shown great promise in representing spatio-temporal and event-driven knowledge. However, existing knowledge graph embedding approaches mainly focus on structural patterns and often overlook the dynamic evolution of entities in both time and space, which limits their effectiveness in downstream reasoning tasks. To address this, we propose a spatio-temporal evolutionary knowledge embedding approach (ST-EKA) that enhances entity representations by modeling their evolution through type-aware encoding, temporal and spatial decay mechanisms, and context aggregation. ST-EKA integrates four core components, including an entity encoder constrained by relational type consistency, a temporal encoder capable of handling both time points and intervals through unified sampling and feedforward encoding, a multi-scale spatial encoder that combines geometric coordinates with semantic attributes, and an evolutionary knowledge encoder that employs attention-based spatio-temporal weighting to capture contextual dynamics. We evaluate ST-EKA on three representative GeoKG datasets—GDELT, ICEWS, and HAD. The results demonstrate that ST-EKA achieves an average improvement of 6.5774% in AUC and 5.0992% in APR on representation learning tasks. In question answering tasks, it yields a maximum average increase of 1.7907% in AUC and 0.5843% in APR. Notably, it exhibits superior performance in chain queries and complex spatio-temporal reasoning, validating its strong robustness, good interpretability, and practical application value. Full article

Food safety regulations and their implementations are becoming increasingly complex due to various reasons such as diverse food sources, supply chain, processing technologies, distribution systems and environmental concerns. Additionally, it is crucial to address diversified consumers and their preferences. To address these multifaceted challenges, adopting an integrated unified management system is essential. This review provides a comprehensive overview of the progressive food safety governance in the Sultanate of Oman. The country is transitioning from a multi-agency to an integrated food control management system. This integrated approach can enhance the coordination between different government agencies and other stakeholders, avoid duplication, identify required resources and ensure optimum use of the resources. The progress can enhance efficiency and effectiveness in managing food safety in Oman. It addresses the issues of the food safety management system, explores the legislative frameworks, risk-based assessment and their enforcement, and creates public awareness and required research for continuous improvement in food safety. This integration approach is expected to continue strengthening food safety governance in the country. Finally, future challenges in achieving food safety are envisioned, including new food sources and technologies, applications of artificial intelligence, and new sensors for quick identification of risks in foods. Full article

The laser damage resistance of optical films holds significant practical importance, as it largely determines both the maximum power output of laser systems and the overall stability of the entire optical assembly. A comprehensive investigation was conducted to examine the influence of both single additives—acetylacetone (ACAC) and diethanolamine (DEA)—and dual-additive systems, specifically ACAC combined with polyethylene glycol 200 (PEG 200) and DEA combined with PEG 200, on TiO₂ film properties and their laser-induced damage behavior under 1064 nm irradiation. It demonstrated that the films fabricated using ACAC exhibited smoother surfaces. Nevertheless, the sol prepared with DEA was more stable, resulting in films with superior optical properties and an enhanced laser-induced damage threshold (LIDT). The incorporation of dual additives further improved the films’ LIDT. Specifically, the film with DEA and PEG 200 achieved the highest LIDT, reaching 21.5 J/cm². Moreover, all films exhibited defect-induced damage, yet distinct damage morphologies were observed across different samples. The single-additive films predominantly displayed stress-type damage patterns, whereas the dual-additive films manifested melting-type damage characteristics. Furthermore, through a combination of experiments and calculations, it was revealed that the reasons why the film with DEA and PEG 200 achieved the highest LIDT were twofold: first, the high thermal conductivity of DEA reduced the maximum temperature at the defect center within the film; second, the long molecular chains of PEG 200 created a looser film structure that better mitigated damage caused by stress and expansion during laser irradiation. This study presents a promising approach to enhancing the LIDT through the strategic selection of additives with high thermal conductivity while simultaneously incorporating organic compounds with long molecular chains to develop effective dual-additive films. Full article

Medical Visual Question Answering (MedVQA) lies at the intersection of computer vision, natural language processing, and clinical decision-making, aiming to generate accurate responses from medical images paired with complex inquiries. Despite recent advances in vision–language models (VLMs), their use in healthcare remains limited by a lack of interpretability and a tendency to produce direct, unexplainable outputs. This opacity undermines their reliability in medical settings, where transparency and justification are critically important. To address this limitation, we propose a zero-shot chain-of-thought prompting framework that guides VLMs to perform multi-step reasoning before arriving at an answer. By encouraging the model to break down the problem, analyze both visual and contextual cues, and construct a stepwise explanation, the approach makes the reasoning process explicit and clinically meaningful. We evaluate the framework on the PMC-VQA benchmark, which includes authentic radiological images and expert-level prompts. In a comparative analysis of three leading VLMs, Gemini 2.5 Pro achieved the highest accuracy (72.48%), followed by Claude 3.5 Sonnet (69.00%) and GPT-4o Mini (67.33%). The results demonstrate that chain-of-thought prompting significantly improves both reasoning transparency and performance in MedVQA tasks. Full article

We investigate the affine term structure models (ATSMs) with unspanned stochastic volatility (USV). Our aim is to test their ability to generate accurate cross-sectional behavior and time-series dynamics of bond yields. Comparing the restricted models and those with USV, we test whether they produce both reasonable estimates for the short rate variance and cross-sectional fit. Essentially, a joint approach from both time series and options data for estimating risk-neutral dynamics in ATSMs should be followed. Due to the scarcity of derivative data in emerging markets, we estimate the model using only time-series of bond yields. A Bayesian estimation approach combining Markov Chain Monte Carlo (MCMC) and the Kalman filter is employed to recover the model parameters and filter out latent state variables. We further incorporate macro-economic indicators and GARCH-based volatility as external validation of the filtered latent volatility process. The $A_1\left(4\right)$USV performs better both in and out of sample, even though the issue of a tension between time series and cross-section remains unresolved. Our findings suggest that even without derivative instruments, it is possible to identify and interpret risk-neutral dynamics and volatility risk using observable time-series data. Full article

Waterway tunnels, a novel type of infrastructure designed for inland waterways in mountainous gorge regions, have seen rapid development in recent years. However, their unique structural characteristics and specific shipping activities pose significant risks in the event of an accident. To enhance the scientific rigor and efficiency of emergency responses to vessel incidents in tunnels, this study focuses on fire accidents in waterway tunnels. Considering the unique challenges of emergency response in such scenarios, we propose an emergency response framework using Business Process Modeling Notation (BPMN). The framework is mapped into a Petri net model encompassing three key stages: detection and early warning, emergency response actions, and recovery. A Colored Hierarchical Timed Petri Net (CHTPN) emergency response model is then developed based on fire incident data and emergency response time functions. Furthermore, a homomorphic Markov chain is employed to assess the network’s validity and performance. Finally, optimization strategies are proposed to improve the emergency response process. The results indicate that the emergency response network demonstrates strong accessibility, effectively mitigating information bottlenecks in critical stages of the response process. The network provides accurate and rapid decision support for different tunnel ship fire scenarios, efficiently and reasonably allocating emergency resources and response teams, and monitoring the operation of key emergency response stages. This enhances the efficiency of emergency operations and provides robust support for decision-making in waterway tunnel fire emergencies. Full article

In the face of economic, social, and environmental pressures, the issue of sustainable development has garnered widespread attention in the Logistics Service Supply Chain (LSSC) with risk attitudes under Technical Output Uncertainty. In this regard, this paper first constructs an optimal emission reduction investment game model for an LSSC composed of Logistics Service Integrators (LSIs) and Logistics Service Providers (LSPs) against the backdrop of Technical Output Uncertainty. To this end, it quantifies the participants’ risk attitudes using a mean-variance model to analyze optimal emission reduction investment decisions for centralized and decentralized LSSC under different levels of risk tolerance. Subsequently, it designs a joint contract with altruistic preferences for sharing emission reduction costs in the LSSC. This contract analyzes the parameter constraints for achieving Pareto optimization within the supply chain. Finally, the study employs a case simulation to analyze the changes in expected revenues for centralized LSSC and joint contracts under different risk tolerance levels. The study reveals that (1) in a centralized LSSC, under risk-neutral attitudes, there exists a unique optimal emission reduction investment, which yields the highest expected return from emission reduction. However, under risk-averse attitudes, the expected return is always lower than the optimal expected return under risk neutrality. (2) In a decentralized LSSC, the emission reduction investment decisions of the Logistics Service Providers are similar to those in a centralized LSSC. (3) Under risk-neutral attitudes, the cost-sharing and altruistic preference-based joint contract can also coordinate the risk-averse LSSC under certain constraints, and by adjusting the cost-sharing and altruistic preference parameters, the expected returns can be reasonably allocated. Full article

Healthy peatlands provide valuable ecosystem services. Peat extraction damages peatlands, leading to carbon emissions. One of the main reasons for peat extraction is for use in horticulture. Replacing peat with recycled organic materials in horticulture is critical to preserve the valuable ecosystems provided by peatlands and to establish more circular supply chains that are reliant on recycling rather than extraction. Despite the strong policy commitment and budget allocation to restore peatlands, the demand for peat-based growing media remains high and drives most of the peat demand. In our research, we mapped the growing media supply chain, held semi-structured interviews with key stakeholders representing different interests, and surveyed amateur gardeners in the UK to understand the bottlenecks experienced by each profile in ending peat use and how to overcome them. We employed semi-structured key expert surveys to understand the supply chain dynamics and consumer demand, informed by these early interviews and the previous literature, we prepared and distributed an online consumer survey and interviewed supply-side stakeholders to understand their perspectives. The findings indicate that the barriers of availability, cost, and performance are shared between the supply-and-demand-side stakeholders. A portfolio of financial, educational and logistic interventions is required to simultaneously support the supply side to accelerate the transformation of production and supply patterns and to aid the demand side to adapt to growing with compost of recycled organic materials. The policies promoting recycled organic material use in horticulture must be coordinated within the UK and with other parts of Europe focusing on the elimination of the peat content in products rather than peat extraction to avoid extraction and the associated destruction of peat stocks elsewhere. Full article

In the context of global climate governance and the low-carbon energy transition, virtual power plant (VPP), a key technology for integrating distributed energy resources, is urgently needed to solve the problem of decentralization and lack of synergy in electricity, carbon, and green certificate trading. Existing studies mostly focus on single energy or carbon trading scenarios and lack a multi-market coupling mechanism supported by blockchain technology, resulting in low transaction transparency and a high risk of information tampering. For this reason, this paper proposes a synergistic optimization strategy for electricity/carbon/green certificate virtual power plants based on blockchain cross-chain transactions. First, Latin Hypercubic Sampling (LHS) is used to generate new energy output and load scenarios, and the K-means clustering method with improved particle swarm optimization are combined to cut down the scenarios and improve the prediction accuracy; second, a relay chain cross-chain trading framework integrating quota system is constructed to realize organic synergy and credible data interaction among electricity, carbon, and green certificate markets; lastly, the multi-energy optimization model of the virtual power plant is designed to integrate carbon capture, Finally, a virtual power plant multi-energy optimization model is designed, integrating carbon capture, power-to-gas (P2G) and other technologies to balance the economy and low-carbon goals. The simulation results show that compared with the traditional model, the proposed strategy reduces the carbon emission intensity by 13.3% (1.43 tons/million CNY), increases the rate of new energy consumption to 98.75%, and partially offsets the cost through the carbon trading revenue, which verifies the Pareto improvement of environmental and economic benefits. This study provides theoretical support for the synergistic optimization of multi-energy markets and helps to build a low-carbon power system with a high proportion of renewable energy. Full article

In knowledge graph construction, missing facts often lead to incomplete structures, thereby limiting the performance of downstream applications. Although recent knowledge graph completion (KGC) methods based on representation learning have achieved notable progress, they still suffer from two fundamental limitations, namely the lack of structured reasoning capabilities and the inability to assess the confidence of their predictions, which often results in unreliable outputs. We propose the GLR framework, which integrates Graph Chain-of-Thought (Graph-CoT) reasoning, LoRA fine-tuning, and the P(True)-based confidence evaluation mechanism. In the KGC task, this approach effectively enhances the reasoning ability and prediction reliability of large language models (LLMs). Specifically, Graph-CoT introduces local subgraph structures to guide LLMs in performing graph-constrained, step-wise reasoning, improving their ability to model multi-hop relational patterns. Complementing this, LoRA-based fine-tuning enables efficient adaptation of LLMs to the KGC scenario with minimal computational overhead, further enhancing the model’s capability for graph-structured reasoning. Moreover, the P(True) mechanism quantifies the reliability of candidate entities, improving the robustness of ranking and the controllability of outputs, thereby enhancing the credibility and interpretability of model predictions in knowledge reasoning tasks. We conducted systematic experiments on the standard KGC datasets FB15K-237, WN18RR, and UMLS, which demonstrate the effectiveness and robustness of the GLR framework. Notably, GLR achieves a Mean Reciprocal Rank (MRR) of 0.507 on FB15K-237, marking a 6.8% improvement over the best recent instruction-tuned method, DIFT combined with CoLE (MRR = 0.439). GLR also maintains significant performance advantages on WN18RR and UMLS, verifying its effectiveness in enhancing both the structured reasoning capabilities and the prediction reliability of LLMs for KGC tasks. These results indicate that GLR offers a unified and scalable solution to enhance structure-aware reasoning and output reliability of LLMs in KGC. Full article

Despite advances in technique and technology, the chainsaw is still the most widely used tool in forestry. For this reason, equipment manufacturers are developing new technical solutions to make working with a chainsaw as easy and efficient as possible. Some examples of this are the development of professional battery-powered chainsaws and the development of new types of saw chains by the leading industry manufacturers. The aim of this paper was to determine the efficiency of the Stihl MSA 300C battery-powered chainsaw equipped with two different types of professional saw chains (Stihl Rapid Super and Stihl Rapid Hexa) when sawing round wood. The efficiency was determined based on measurements of electricity consumption, sawing speed, sawn wood cross-section, and wood chips and dust mass produced during sawing. The second aim was to determine whether there is a difference in measured vibration magnitude between the two tested saw chains. Fresh-fallen European beech (Fagus sylvatica L.) log, approx. 25 cm diameter without pronounced ellipticity, was used for sampling. Results indicate that although the saw chain manufacturer claims the new type of saw chain (Stihl Rapid Hexa) enables greater efficiency of the chainsaw, this was not the case. Results point to a 37% increase in mean sawing time, as well as a 23% increase in energy consumption when using the Rapid Hexa chain, with statistically significant difference (p ≤ 0.05). It should be emphasized that the manual operation of the chainsaw does not allow for a reliable determination of differences in energy consumption caused by changes in saw chain geometry. The advantages of this saw chain are that it is easier to maintain (sharpen) and significantly less wood chips and dust are produced. The measured vibration magnitude shows a statistically significant difference (p ≤ 0.05), i.e., a lower vibration total value on the front handle when using the Stihl Rapid Hexa chain. Full article