The reliable operation of power grid secondary equipment is an important guarantee for the safety and stability of the power system.However,various defects could be produced in the secondary equipment during longtermo...The reliable operation of power grid secondary equipment is an important guarantee for the safety and stability of the power system.However,various defects could be produced in the secondary equipment during longtermoperation.The complex relationship between the defect phenomenon andmulti-layer causes and the probabilistic influence of secondary equipment cannot be described through knowledge extraction and fusion technology by existing methods,which limits the real-time and accuracy of defect identification.Therefore,a defect recognition method based on the Bayesian network and knowledge graph fusion is proposed.The defect data of secondary equipment is transformed into the structured knowledge graph through knowledge extraction and fusion technology.The knowledge graph of power grid secondary equipment is mapped to the Bayesian network framework,combined with historical defect data,and introduced Noisy-OR nodes.The prior and conditional probabilities of the Bayesian network are then reasonably assigned to build a model that reflects the probability dependence between defect phenomena and potential causes in power grid secondary equipment.Defect identification of power grid secondary equipment is achieved by defect subgraph search based on the knowledge graph,and defect inference based on the Bayesian network.Practical application cases prove this method’s effectiveness in identifying secondary equipment defect causes,improving identification accuracy and efficiency.展开更多
Since Google introduced the concept of Knowledge Graphs(KGs)in 2012,their construction technologies have evolved into a comprehensive methodological framework encompassing knowledge acquisition,extraction,representati...Since Google introduced the concept of Knowledge Graphs(KGs)in 2012,their construction technologies have evolved into a comprehensive methodological framework encompassing knowledge acquisition,extraction,representation,modeling,fusion,computation,and storage.Within this framework,knowledge extraction,as the core component,directly determines KG quality.In military domains,traditional manual curation models face efficiency constraints due to data fragmentation,complex knowledge architectures,and confidentiality protocols.Meanwhile,crowdsourced ontology construction approaches from general domains prove non-transferable,while human-crafted ontologies struggle with generalization deficiencies.To address these challenges,this study proposes an OntologyAware LLM Methodology for Military Domain Knowledge Extraction(LLM-KE).This approach leverages the deep semantic comprehension capabilities of Large Language Models(LLMs)to simulate human experts’cognitive processes in crowdsourced ontology construction,enabling automated extraction of military textual knowledge.It concurrently enhances knowledge processing efficiency and improves KG completeness.Empirical analysis demonstrates that this method effectively resolves scalability and dynamic adaptation challenges in military KG construction,establishing a novel technological pathway for advancing military intelligence development.展开更多
The rapid development of information technology and accelerated digitalization have led to an explosive growth of data across various fields.As a key technology for knowledge representation and sharing,knowledge graph...The rapid development of information technology and accelerated digitalization have led to an explosive growth of data across various fields.As a key technology for knowledge representation and sharing,knowledge graphs play a crucial role by constructing structured networks of relationships among entities.However,data sparsity and numerous unexplored implicit relations result in the widespread incompleteness of knowledge graphs.In static knowledge graph completion,most existing methods rely on linear operations or simple interaction mechanisms for triple encoding,making it difficult to fully capture the deep semantic associations between entities and relations.Moreover,many methods focus only on the local information of individual triples,ignoring the rich semantic dependencies embedded in the neighboring nodes of entities within the graph structure,which leads to incomplete embedding representations.To address these challenges,we propose Two-Stage Mixer Embedding(TSMixerE),a static knowledge graph completion method based on entity context.In the unit semantic extraction stage,TSMixerE leveragesmulti-scale circular convolution to capture local features atmultiple granularities,enhancing the flexibility and robustness of feature interactions.A channel attention mechanism amplifies key channel responses to suppress noise and irrelevant information,thereby improving the discriminative power and semantic depth of feature representations.For contextual information fusion,a multi-layer self-attentionmechanism enables deep interactions among contextual cues,effectively integrating local details with global context.Simultaneously,type embeddings clarify the semantic identities and roles of each component,enhancing the model’s sensitivity and fusion capabilities for diverse information sources.Furthermore,TSMixerE constructs contextual unit sequences for entities,fully exploring neighborhood information within the graph structure to model complex semantic dependencies,thus improving the completeness and generalization of embedding representations.展开更多
Active fault-tolerant control utilizes information obtained from fault diagnosis to reconfigure the control law to compensate for faults in the wastewater treatment process. However, since the similarity of fault char...Active fault-tolerant control utilizes information obtained from fault diagnosis to reconfigure the control law to compensate for faults in the wastewater treatment process. However, since the similarity of fault characteristic in the incipient stage can result in misdiagnosis, it is a challenge for fault-tolerant control to ensure system safety and reliability. Therefore, to address this issue, a fault diagnosis and fault-tolerant control with a knowledge transfer strategy(KT-FDFTC) is proposed in this paper. First, a knowledge reasoning diagnosis strategy using multi-source transfer learning is designed to distinguish the similar characteristic of incipient faults. Then, the multi-source knowledge can assist in the diagnosis strategy to strengthen the fault information for fault-tolerant control. Second, a knowledge adaptive compensation mechanism, which makes knowledge and data coupled into the output trajectory regarded as an objective function, is employed to dynamically compute the control law. Then, KT-FDFTC can ensure the stable operation to adapt to various fault conditions. Third, the Lyapunov function is established to demonstrate the stability of KT-FDFTC. Then, the theoretical basis can offer the successful application of KTFDFTC. Finally, the proposed method is validated through a real WWTP and a simulation platform. The experimental results confirm that KT-FDFTC can provide good diagnosis performance and fault tolerance ability.展开更多
This paper delves into effective pathways for transforming course ecosystems from resource provision to knowledge service and competency development through university-enterprise collaboration in co-building knowledge...This paper delves into effective pathways for transforming course ecosystems from resource provision to knowledge service and competency development through university-enterprise collaboration in co-building knowledge graphs and intelligent shared courses.This approach enables personalized,learning-driven teaching.Based on knowledge graphs and integrated teacher-machine-student smart teaching scenarios,it not only innovates autonomous learning environments and human-computer interaction models while optimizing teaching experiences for both instructors and students,but also effectively addresses the issues of students’“scattered,superficial,and fragmented learning”.This establishes the foundation for personalized teaching tailored to individual aptitudes.展开更多
This paper proposes a deep learning-based 3D LiDAR perception framework designed for applications such as autonomous robots and vehicles.To address the high dependency on large-scale annotated data—an inherent limita...This paper proposes a deep learning-based 3D LiDAR perception framework designed for applications such as autonomous robots and vehicles.To address the high dependency on large-scale annotated data—an inherent limitation of deep learning models—this study introduces a hybrid perception architecture that incorporates expertdriven LiDAR processing techniques into the deep neural network.Traditional 3DLiDAR processingmethods typically remove ground planes and apply distance-or density-based clustering for object detection.In this work,such expert knowledge is encoded as feature-level inputs and fused with the deep network,therebymitigating the data dependency issue of conventional learning-based approaches.Specifically,the proposedmethod combines two expert algorithms—Patchwork++for ground segmentation and DBSCAN for clustering—with a PointPillars-based LiDAR detection network.We design four hybrid versions of the network depending on the stage and method of integrating expert features into the feature map of the deep model.Among these,Version 4 incorporates a modified neck structure in PointPillars and introduces a new Cluster 2D Pseudo-Map Branch that utilizes cluster-level pseudo-images generated from Patchwork++and DBSCAN.This version achieved a+3.88%improvement mean Average Precision(mAP)compared to the baseline PointPillars.The results demonstrate that embedding expert-based perception logic into deep neural architectures can effectively enhance performance and reduce dependency on extensive training datasets,offering a promising direction for robust 3D LiDAR object detection in real-world scenarios.展开更多
Wetting deformation in earth-rockfill dams is a critical factor influencingdam safety.Although numerous mathematical models have been developed to describe this phenomenon,most of them rely on empirical formulations a...Wetting deformation in earth-rockfill dams is a critical factor influencingdam safety.Although numerous mathematical models have been developed to describe this phenomenon,most of them rely on empirical formulations and lack prior knowledge of model parameters,which is essential for Bayesian parameter inversion to enhance accuracy and reduce uncertainty.This study introduces a datadriven approach to establishing prior knowledge of earth-rockfill dams.Driving factors are utilized to determine the potential range of model parameters,and settlement changes within this range are calculated.The results are iteratively compared with actual monitoring data until the calculated range encompasses the observed data,thereby providing prior knowledge of the model parameters.The proposed method is applied to the right-bank earth-rockfilldam of Danjiangkou.Employing a Gibbs sample size of 30,000,the proposed method effectively calibrates the prior knowledge of the wetting model parameters,achieving a root mean square error(RMSE)of 5.18 mm for the settlement predictions.By comparison,the use of non-informative priors with sample sizes of 30,000 and 50,000 results in significantly larger RMSE values of 11.97 mm and 16.07 mm,respectively.Furthermore,the computational efficiencyof the proposed method is demonstrated by an inversion computation time of 902 s for 30,000 samples,which is notably shorter than the 1026 s and 1558 s required for noninformative priors with 30,000 and 50,000 samples,respectively.These findingsunderscore the superior performance of the proposed approach in terms of both prediction accuracy and computational efficiency.These results demonstrate that the proposed method not only improves the predictive accuracy but also enhances the computational efficiency,enabling optimal parameter identificationwith reduced computational effort.This approach provides a robust and efficientframework for advancing dam safety assessments.展开更多
The viscosity of refining slags plays a critical role in metallurgical processes.However,obtaining accurate viscosity data remains challenging due to the complexities of high-temperature experiments,often relying on e...The viscosity of refining slags plays a critical role in metallurgical processes.However,obtaining accurate viscosity data remains challenging due to the complexities of high-temperature experiments,often relying on empirical models with limited predictive capabilities.This study focuses on the influence of optical basicity on viscosity in CaO-Al_(2)O_(3)-based refining slags,leveraging machine learning to address data scarcity and improve prediction accuracy.An automated framework for algorithm integration,parameter tuning,and evaluation ranking framework(Auto-APE)is employed to develop customized data-driven models for various slag systems,including CaO-Al_(2)O_(3)-SiO_(2),CaO-Al_(2)O_(3)-CaF_(2),CaO-Al_(2)O_(3)-SiO_(2)-MgO,and CaO-Al_(2)O_(3)-SiO_(2)-MgO-CaF_(2).By incorporating optical basicity as a key feature,the models achieve an average validation error of 8.0%to 15.1%,significantly outperforming traditional empirical models.Additionally,symbolic regression is introduced to rapidly construct domain-specific features,such as optical basicity-like descriptors,offering a potential breakthrough in performance prediction for small datasets.This work highlights the critical role of domain-specific knowledge in understanding and predicting viscosity,providing a robust machine learning-based approach for optimizing refining slag properties.展开更多
Aiming at the problem of insufficient recognition of implicit variants by existing Chinese sensitive text detection methods,this paper proposes the IPKE-MoE framework,which consists of three parts,namely,a sensitive w...Aiming at the problem of insufficient recognition of implicit variants by existing Chinese sensitive text detection methods,this paper proposes the IPKE-MoE framework,which consists of three parts,namely,a sensitive word variant extraction framework,a sensitive word variant knowledge enhancement layer and a mixture-of-experts(MoE)classification layer.First,sensitive word variants are precisely extracted through dynamic iterative prompt templates and the context-aware capabilities of Large Language Models(LLMs).Next,the extracted variants are used to construct a knowledge enhancement layer for sensitive word variants based on RoCBert models.Specifically,after locating variants via n-gram algorithms,variant types are mapped to embedding vectors and fused with original word vectors.Finally,a mixture-of-experts(MoE)classification layer is designed(sensitive word,sentiment,and semantic experts),which decouples the relationship between sensitiveword existence and text toxicity throughmultiple experts.This framework effectively combines the comprehension ability of Large Language Models(LLMs)with the discriminative ability of smaller models.Our two experiments demonstrate that the sensitive word variant extraction framework based on dynamically iterated prompt templates outperforms other baseline prompt templates.TheRoCBert models incorporating the sensitive word variant knowledge enhancement layer and a mixture-of-experts(MoE)classification layer achieve superior classification performance compared to other baselines.展开更多
With the widespread use of social media,the propagation of health-related rumors has become a significant public health threat.Existing methods for detecting health rumors predominantly rely on external knowledge or p...With the widespread use of social media,the propagation of health-related rumors has become a significant public health threat.Existing methods for detecting health rumors predominantly rely on external knowledge or propagation structures,with only a few recent approaches attempting causal inference;however,these have not yet effectively integrated causal discovery with domain-specific knowledge graphs for detecting health rumors.In this study,we found that the combined use of causal discovery and domain-specific knowledge graphs can effectively identify implicit pseudo-causal logic embedded within texts,holding significant potential for health rumor detection.To this end,we propose CKDG—a dual-graph fusion framework based on causal logic and medical knowledge graphs.CKDG constructs a weighted causal graph to capture the implicit causal relationships in the text and introduces a medical knowledge graph to verify semantic consistency,thereby enhancing the ability to identify the misuse of professional terminology and pseudoscientific claims.In experiments conducted on a dataset comprising 8430 health rumors,CKDG achieved an accuracy of 91.28%and an F1 score of 90.38%,representing improvements of 5.11%and 3.29%over the best baseline,respectively.Our results indicate that the integrated use of causal discovery and domainspecific knowledge graphs offers significant advantages for health rumor detection systems.This method not only improves detection performance but also enhances the transparency and credibility of model decisions by tracing causal chains and sources of knowledge conflicts.We anticipate that this work will provide key technological support for the development of trustworthy health-information filtering systems,thereby improving the reliability of public health information on social media.展开更多
Knowledge distillation has become a standard technique for compressing large language models into efficient student models,but existing methods often struggle to balance prediction accuracy with explanation quality.Re...Knowledge distillation has become a standard technique for compressing large language models into efficient student models,but existing methods often struggle to balance prediction accuracy with explanation quality.Recent approaches such as Distilling Step-by-Step(DSbS)introduce explanation supervision,yet they apply it in a uniform manner that may not fully exploit the different learning dynamics of prediction and explanation.In this work,we propose a task-structured curriculum learning(TSCL)framework that structures training into three sequential phases:(i)prediction-only,to establish stable feature representations;(ii)joint prediction-explanation,to align task outputs with rationale generation;and(iii)explanation-only,to refine the quality of rationales.This design provides a simple but effective modification to DSbS,requiring no architectural changes and adding negligible training cost.We justify the phase scheduling with ablation studies and convergence analysis,showing that an initial prediction-heavy stage followed by a balanced joint phase improves both stability and explanation alignment.Extensive experiments on five datasets(e-SNLI,ANLI,CommonsenseQA,SVAMP,and MedNLI)demonstrate that TSCL consistently outperforms strong baselines,achieving gains of+1.7-2.6 points in accuracy and 0.8-1.2 in ROUGE-L,corresponding to relative error reductions of up to 21%.Beyond lexical metrics,human evaluation and ERASERstyle faithfulness diagnostics confirm that TSCL produces more faithful and informative explanations.Comparative training curves further reveal faster convergence and lower variance across seeds.Efficiency analysis shows less than 3%overhead in wall-clock training time and no additional inference cost,making the approach practical for realworld deployment.This study demonstrates that a simple task-structured curriculum can significantly improve the effectiveness of knowledge distillation.By separating and sequencing objectives,TSCL achieves a better balance between accuracy,stability,and explanation quality.The framework generalizes across domains,including medical NLI,and offers a principled recipe for future applications in multimodal reasoning and reinforcement learning.展开更多
With the large-scale deployment of the Internet ofThings(IoT)devices,their weak securitymechanisms make them prime targets for malware attacks.Attackers often use Domain Generation Algorithm(DGA)to generate random dom...With the large-scale deployment of the Internet ofThings(IoT)devices,their weak securitymechanisms make them prime targets for malware attacks.Attackers often use Domain Generation Algorithm(DGA)to generate random domain names,hiding the real IP of Command and Control(C&C)servers to build botnets.Due to the randomness and dynamics of DGA,traditional methods struggle to detect them accurately,increasing the difficulty of network defense.This paper proposes a lightweight DGA detection model based on knowledge distillation for resource-constrained IoT environments.Specifically,a teacher model combining CharacterBERT,a bidirectional long short-term memory(BiLSTM)network,and attention mechanism(ATT)is constructed:it extracts character-level semantic features viaCharacterBERT,captures sequence dependencieswith the BiLSTM,and integrates theATT for key feature weighting,formingmulti-granularity feature fusion.An improved knowledge distillation approach transfers the teacher model’s learned knowledge to the simplified DistilBERT student model.Experimental results show the teacher model achieves 98.68%detection accuracy.The student modelmaintains slightly improved accuracy while significantly compressing parameters to approximately 38.4%of the teacher model’s scale,greatly reducing computational overhead for IoT deployment.展开更多
Amidst evolving user behavior driven by the development of the internet,enhancing the operational quality of trade publishing knowledge service platforms has become a significant challenge for publishing institutions....Amidst evolving user behavior driven by the development of the internet,enhancing the operational quality of trade publishing knowledge service platforms has become a significant challenge for publishing institutions.To address this issue,this paper employs a combined approach of theoretical analysis and case study,introducing the SICAS(Sense-Interest-Connection-Action-Share)user consumption behavior analysis model and selecting“CITIC Academy”as the case study subject.It systematically examines and summarizes the platform’s operational practices and specific strategies,aiming to offer strategic insights and practical references for the operational improvement and sustainable,high-quality development of trade publishing knowledge service platforms.展开更多
Conversational recommender systems(CRSs)focus on refining preferences and providing personalized recommendations through natural language interactions and dialogue history.Large language models(LLMs)have shown outstan...Conversational recommender systems(CRSs)focus on refining preferences and providing personalized recommendations through natural language interactions and dialogue history.Large language models(LLMs)have shown outstanding performance across various domains,thereby prompting researchers to investigate their applicability in recommendation systems.However,due to the lack of task-specific knowledge and an inefficient feature extraction process,LLMs still have suboptimal performance in recommendation tasks.Therefore,external knowledge sources,such as knowledge graphs(KGs)and knowledge bases(KBs),are often introduced to address the issue of data sparsity.Compared to KGs,KBs possess higher retrieval efficiency,making them more suitable for scenarios where LLMs serve as recommenders.To this end,we introduce a novel framework integrating LLMs with KBs for enhanced retrieval generation,namely LLMKB.LLMKB initially leverages structured knowledge to create mapping dictionaries,extracting entity-relation information from heterogeneous knowledge to construct KBs.Then,LLMKB achieves the embedding calibration between user information representations and documents in KBs through retrieval model fine-tuning.Finally,LLMKB employs retrievalaugmented generation to produce recommendations based on fused text inputs,followed by post-processing.Experiment results on two public CRS datasets demonstrate the effectiveness of our framework.Our code is publicly available at the link:https://anonymous.4open.science/r/LLMKB-6FD0.展开更多
Bird's-eye-view(BEV)perception is a core technology for autonomous driving systems.However,existing solutions face the dilemma of high costs associated with multimodal methods and limited performance of vision-onl...Bird's-eye-view(BEV)perception is a core technology for autonomous driving systems.However,existing solutions face the dilemma of high costs associated with multimodal methods and limited performance of vision-only approaches.To address this issue,this paper proposes a framework named“a lightweight pure visual BEV perception method based on dual distillation of spatial-temporal knowledge”.This framework innovatively designs a lightweight vision-only student model based on Res Net,which leverages a dual distillation mechanism to learn from a powerful teacher model that integrates temporal information from both image and light detection and ranging(LiDAR)modalities.Specifically,we distill efficient multi-modal feature extraction and spatial fusion capabilities from the BEVFusion model,and distill advanced temporal information fusion and spatiotemporal attention mechanisms from the BEVFormer model.This dual distillation strategy enables the student model to achieve perception performance close to that of multi-modal models without relying on Li DAR.Experimental results on the nu Scenes dataset demonstrate that the proposed model significantly outperforms classical vision-only algorithms,achieves comparable performance to current state-of-the-art vision-only methods on the nu Scenes detection leaderboard in terms of both mean average precision(mAP)and the nu Scenes detection score(NDS)metrics,and exhibits notable advantages in inference computational efficiency.Although the proposed dual-teacher paradigm incurs higher offline training costs compared to single-model approaches,it yields a streamlined and highly efficient student model suitable for resource-constrained real-time deployment.This provides an effective pathway toward low-cost,high-performance autonomous driving perception systems.展开更多
Large language models(LLMs)have significantly advanced artificial intelligence(AI)by excelling in tasks such as understanding,generation,and reasoning across multiple modalities.Despite these achievements,LLMs have in...Large language models(LLMs)have significantly advanced artificial intelligence(AI)by excelling in tasks such as understanding,generation,and reasoning across multiple modalities.Despite these achievements,LLMs have inherent limitations including outdated information,hallucinations,inefficiency,lack of interpretability,and challenges in domain-specific accuracy.To address these issues,this survey explores three promising directions in the post-LLM era:knowledge empowerment,model collaboration,and model co-evolution.First,we examine methods of integrating external knowledge into LLMs to enhance factual accuracy,reasoning capabilities,and interpretability,including incorporating knowledge into training objectives,instruction tuning,retrieval-augmented inference,and knowledge prompting.Second,we discuss model collaboration strategies that leverage the complementary strengths of LLMs and smaller models to improve efficiency and domain-specific performance through techniques such as model merging,functional model collaboration,and knowledge injection.Third,we delve into model co-evolution,in which multiple models collaboratively evolve by sharing knowledge,parameters,and learning strategies to adapt to dynamic environments and tasks,thereby enhancing their adaptability and continual learning.We illustrate how the integration of these techniques advances AI capabilities in science,engineering,and society—particularly in hypothesis development,problem formulation,problem-solving,and interpretability across various domains.We conclude by outlining future pathways for further advancement and applications.展开更多
To further improve the level and quality of art design education in colleges and universities and promote the reform and innovation of higher education,this paper conducts research from the perspective of intangible c...To further improve the level and quality of art design education in colleges and universities and promote the reform and innovation of higher education,this paper conducts research from the perspective of intangible cultural heritage(ICH)knowledge graph empowering education,aiming to break through the current bottlenecks in the integration of ICH and art education and advance the innovative reform of art design education in colleges and universities.Firstly,it analyzes and interprets the significance of ICH knowledge graph empowering college art design education.Then,it elaborates on the challenges faced in this process.Finally,it proposes multiple educational measures,including optimizing graph construction,strengthening literacy education,deepening curriculum integration,and improving the support system,so as to effectively promote the innovative reform of college art design education and enhance the level and quality of talent cultivation in colleges and universities.展开更多
Knowledge graphs(KGs)have been widely accepted as powerful tools for modeling the complex relationships between concepts and developing knowledge-based services.In recent years,researchers in the field of power system...Knowledge graphs(KGs)have been widely accepted as powerful tools for modeling the complex relationships between concepts and developing knowledge-based services.In recent years,researchers in the field of power systems have explored KGs to develop intelligent dispatching systems for increasingly large power grids.With multiple power grid dispatching knowledge graphs(PDKGs)constructed by different agencies,the knowledge fusion of different PDKGs is useful for providing more accurate decision supports.To achieve this,entity alignment that aims at connecting different KGs by identifying equivalent entities is a critical step.Existing entity alignment methods cannot integrate useful structural,attribute,and relational information while calculating entities’similarities and are prone to making many-to-one alignments,thus can hardly achieve the best performance.To address these issues,this paper proposes a collective entity alignment model that integrates three kinds of available information and makes collective counterpart assignments.This model proposes a novel knowledge graph attention network(KGAT)to learn the embeddings of entities and relations explicitly and calculates entities’similarities by adaptively incorporating the structural,attribute,and relational similarities.Then,we formulate the counterpart assignment task as an integer programming(IP)problem to obtain one-to-one alignments.We not only conduct experiments on a pair of PDKGs but also evaluate o ur model on three commonly used cross-lingual KGs.Experimental comparisons indicate that our model outperforms other methods and provides an effective tool for the knowledge fusion of PDKGs.展开更多
In the international shipping industry, digital intelligence transformation has become essential, with both governments and enterprises actively working to integrate diverse datasets. The domain of maritime and shippi...In the international shipping industry, digital intelligence transformation has become essential, with both governments and enterprises actively working to integrate diverse datasets. The domain of maritime and shipping is characterized by a vast array of document types, filled with complex, large-scale, and often chaotic knowledge and relationships. Effectively managing these documents is crucial for developing a Large Language Model (LLM) in the maritime domain, enabling practitioners to access and leverage valuable information. A Knowledge Graph (KG) offers a state-of-the-art solution for enhancing knowledge retrieval, providing more accurate responses and enabling context-aware reasoning. This paper presents a framework for utilizing maritime and shipping documents to construct a knowledge graph using GraphRAG, a hybrid tool combining graph-based retrieval and generation capabilities. The extraction of entities and relationships from these documents and the KG construction process are detailed. Furthermore, the KG is integrated with an LLM to develop a Q&A system, demonstrating that the system significantly improves answer accuracy compared to traditional LLMs. Additionally, the KG construction process is up to 50% faster than conventional LLM-based approaches, underscoring the efficiency of our method. This study provides a promising approach to digital intelligence in shipping, advancing knowledge accessibility and decision-making.展开更多
In the context of power generation companies, vast amounts of specialized data and expert knowledge have been accumulated. However, challenges such as data silos and fragmented knowledge hinder the effective utilizati...In the context of power generation companies, vast amounts of specialized data and expert knowledge have been accumulated. However, challenges such as data silos and fragmented knowledge hinder the effective utilization of this information. This study proposes a novel framework for intelligent Question-and-Answer (Q&A) systems based on Retrieval-Augmented Generation (RAG) to address these issues. The system efficiently acquires domain-specific knowledge by leveraging external databases, including Relational Databases (RDBs) and graph databases, without additional fine-tuning for Large Language Models (LLMs). Crucially, the framework integrates a Dynamic Knowledge Base Updating Mechanism (DKBUM) and a Weighted Context-Aware Similarity (WCAS) method to enhance retrieval accuracy and mitigate inherent limitations of LLMs, such as hallucinations and lack of specialization. Additionally, the proposed DKBUM dynamically adjusts knowledge weights within the database, ensuring that the most recent and relevant information is utilized, while WCAS refines the alignment between queries and knowledge items by enhanced context understanding. Experimental validation demonstrates that the system can generate timely, accurate, and context-sensitive responses, making it a robust solution for managing complex business logic in specialized industries.展开更多
基金supported by the State Grid Southwest Branch Project“Research on Defect Diagnosis and Early Warning Technology of Relay Protection and Safety Automation Devices Based on Multi-Source Heterogeneous Defect Data”.
文摘The reliable operation of power grid secondary equipment is an important guarantee for the safety and stability of the power system.However,various defects could be produced in the secondary equipment during longtermoperation.The complex relationship between the defect phenomenon andmulti-layer causes and the probabilistic influence of secondary equipment cannot be described through knowledge extraction and fusion technology by existing methods,which limits the real-time and accuracy of defect identification.Therefore,a defect recognition method based on the Bayesian network and knowledge graph fusion is proposed.The defect data of secondary equipment is transformed into the structured knowledge graph through knowledge extraction and fusion technology.The knowledge graph of power grid secondary equipment is mapped to the Bayesian network framework,combined with historical defect data,and introduced Noisy-OR nodes.The prior and conditional probabilities of the Bayesian network are then reasonably assigned to build a model that reflects the probability dependence between defect phenomena and potential causes in power grid secondary equipment.Defect identification of power grid secondary equipment is achieved by defect subgraph search based on the knowledge graph,and defect inference based on the Bayesian network.Practical application cases prove this method’s effectiveness in identifying secondary equipment defect causes,improving identification accuracy and efficiency.
文摘Since Google introduced the concept of Knowledge Graphs(KGs)in 2012,their construction technologies have evolved into a comprehensive methodological framework encompassing knowledge acquisition,extraction,representation,modeling,fusion,computation,and storage.Within this framework,knowledge extraction,as the core component,directly determines KG quality.In military domains,traditional manual curation models face efficiency constraints due to data fragmentation,complex knowledge architectures,and confidentiality protocols.Meanwhile,crowdsourced ontology construction approaches from general domains prove non-transferable,while human-crafted ontologies struggle with generalization deficiencies.To address these challenges,this study proposes an OntologyAware LLM Methodology for Military Domain Knowledge Extraction(LLM-KE).This approach leverages the deep semantic comprehension capabilities of Large Language Models(LLMs)to simulate human experts’cognitive processes in crowdsourced ontology construction,enabling automated extraction of military textual knowledge.It concurrently enhances knowledge processing efficiency and improves KG completeness.Empirical analysis demonstrates that this method effectively resolves scalability and dynamic adaptation challenges in military KG construction,establishing a novel technological pathway for advancing military intelligence development.
基金supported by the National Natural Science Foundation of China(No.62267005)the Chinese Guangxi Natural Science Foundation(No.2023GXNSFAA026493)+1 种基金Guangxi Collaborative Innovation Center ofMulti-Source Information Integration and Intelligent ProcessingGuangxi Academy of Artificial Intelligence.
文摘The rapid development of information technology and accelerated digitalization have led to an explosive growth of data across various fields.As a key technology for knowledge representation and sharing,knowledge graphs play a crucial role by constructing structured networks of relationships among entities.However,data sparsity and numerous unexplored implicit relations result in the widespread incompleteness of knowledge graphs.In static knowledge graph completion,most existing methods rely on linear operations or simple interaction mechanisms for triple encoding,making it difficult to fully capture the deep semantic associations between entities and relations.Moreover,many methods focus only on the local information of individual triples,ignoring the rich semantic dependencies embedded in the neighboring nodes of entities within the graph structure,which leads to incomplete embedding representations.To address these challenges,we propose Two-Stage Mixer Embedding(TSMixerE),a static knowledge graph completion method based on entity context.In the unit semantic extraction stage,TSMixerE leveragesmulti-scale circular convolution to capture local features atmultiple granularities,enhancing the flexibility and robustness of feature interactions.A channel attention mechanism amplifies key channel responses to suppress noise and irrelevant information,thereby improving the discriminative power and semantic depth of feature representations.For contextual information fusion,a multi-layer self-attentionmechanism enables deep interactions among contextual cues,effectively integrating local details with global context.Simultaneously,type embeddings clarify the semantic identities and roles of each component,enhancing the model’s sensitivity and fusion capabilities for diverse information sources.Furthermore,TSMixerE constructs contextual unit sequences for entities,fully exploring neighborhood information within the graph structure to model complex semantic dependencies,thus improving the completeness and generalization of embedding representations.
基金supported by the National Natural Science Foundation of China (Grant Nos.62125301,62021003,62303024,U24A20275,62522302,62473011,92467205)the National Key Research and Development Project (Grant Nos.2022YFB3305800-5,2024YFE0212400)+2 种基金the Youth Beijing Scholars Program (Grant No.037)the Beijing Nova Program (Grant Nos.20240484694,20250484938)the Beijing Natural Science Foundation (Grant No.L253010)。
文摘Active fault-tolerant control utilizes information obtained from fault diagnosis to reconfigure the control law to compensate for faults in the wastewater treatment process. However, since the similarity of fault characteristic in the incipient stage can result in misdiagnosis, it is a challenge for fault-tolerant control to ensure system safety and reliability. Therefore, to address this issue, a fault diagnosis and fault-tolerant control with a knowledge transfer strategy(KT-FDFTC) is proposed in this paper. First, a knowledge reasoning diagnosis strategy using multi-source transfer learning is designed to distinguish the similar characteristic of incipient faults. Then, the multi-source knowledge can assist in the diagnosis strategy to strengthen the fault information for fault-tolerant control. Second, a knowledge adaptive compensation mechanism, which makes knowledge and data coupled into the output trajectory regarded as an objective function, is employed to dynamically compute the control law. Then, KT-FDFTC can ensure the stable operation to adapt to various fault conditions. Third, the Lyapunov function is established to demonstrate the stability of KT-FDFTC. Then, the theoretical basis can offer the successful application of KTFDFTC. Finally, the proposed method is validated through a real WWTP and a simulation platform. The experimental results confirm that KT-FDFTC can provide good diagnosis performance and fault tolerance ability.
基金supported by Harbin Institute of Technology High-level Teaching Achievement Award(National Level)Cultivation Project(256709).
文摘This paper delves into effective pathways for transforming course ecosystems from resource provision to knowledge service and competency development through university-enterprise collaboration in co-building knowledge graphs and intelligent shared courses.This approach enables personalized,learning-driven teaching.Based on knowledge graphs and integrated teacher-machine-student smart teaching scenarios,it not only innovates autonomous learning environments and human-computer interaction models while optimizing teaching experiences for both instructors and students,but also effectively addresses the issues of students’“scattered,superficial,and fragmented learning”.This establishes the foundation for personalized teaching tailored to individual aptitudes.
基金supported by Basic Science Research Program through the National Research Foundation of Korea(NRF)funded by the Ministry of Education(RS-2023-00245084)by Korea Institute for Advancement of Technology(KIAT)grant funded by the Korea Government(MOTIE)(RS-2024-00415938,HRD Program for Industrial Innovation)and Soonchunhyang University.
文摘This paper proposes a deep learning-based 3D LiDAR perception framework designed for applications such as autonomous robots and vehicles.To address the high dependency on large-scale annotated data—an inherent limitation of deep learning models—this study introduces a hybrid perception architecture that incorporates expertdriven LiDAR processing techniques into the deep neural network.Traditional 3DLiDAR processingmethods typically remove ground planes and apply distance-or density-based clustering for object detection.In this work,such expert knowledge is encoded as feature-level inputs and fused with the deep network,therebymitigating the data dependency issue of conventional learning-based approaches.Specifically,the proposedmethod combines two expert algorithms—Patchwork++for ground segmentation and DBSCAN for clustering—with a PointPillars-based LiDAR detection network.We design four hybrid versions of the network depending on the stage and method of integrating expert features into the feature map of the deep model.Among these,Version 4 incorporates a modified neck structure in PointPillars and introduces a new Cluster 2D Pseudo-Map Branch that utilizes cluster-level pseudo-images generated from Patchwork++and DBSCAN.This version achieved a+3.88%improvement mean Average Precision(mAP)compared to the baseline PointPillars.The results demonstrate that embedding expert-based perception logic into deep neural architectures can effectively enhance performance and reduce dependency on extensive training datasets,offering a promising direction for robust 3D LiDAR object detection in real-world scenarios.
基金supported by the National Key R&D Program of China(Grant No.2023YFC3209504)Natural Science Foundation of Wuhan(Grant No.2024040801020271)the Fundamental Research Funds for Central Public Welfare Research Institutes(Grant No.CKSF2025718/YT).
文摘Wetting deformation in earth-rockfill dams is a critical factor influencingdam safety.Although numerous mathematical models have been developed to describe this phenomenon,most of them rely on empirical formulations and lack prior knowledge of model parameters,which is essential for Bayesian parameter inversion to enhance accuracy and reduce uncertainty.This study introduces a datadriven approach to establishing prior knowledge of earth-rockfill dams.Driving factors are utilized to determine the potential range of model parameters,and settlement changes within this range are calculated.The results are iteratively compared with actual monitoring data until the calculated range encompasses the observed data,thereby providing prior knowledge of the model parameters.The proposed method is applied to the right-bank earth-rockfilldam of Danjiangkou.Employing a Gibbs sample size of 30,000,the proposed method effectively calibrates the prior knowledge of the wetting model parameters,achieving a root mean square error(RMSE)of 5.18 mm for the settlement predictions.By comparison,the use of non-informative priors with sample sizes of 30,000 and 50,000 results in significantly larger RMSE values of 11.97 mm and 16.07 mm,respectively.Furthermore,the computational efficiencyof the proposed method is demonstrated by an inversion computation time of 902 s for 30,000 samples,which is notably shorter than the 1026 s and 1558 s required for noninformative priors with 30,000 and 50,000 samples,respectively.These findingsunderscore the superior performance of the proposed approach in terms of both prediction accuracy and computational efficiency.These results demonstrate that the proposed method not only improves the predictive accuracy but also enhances the computational efficiency,enabling optimal parameter identificationwith reduced computational effort.This approach provides a robust and efficientframework for advancing dam safety assessments.
基金supported by the National Key Research and Development Program of China(No.2023YFB3712401),the National Natural Science Foundation of China(No.52274301)the Aeronautical Science Foundation of China(No.2023Z0530S6005)the Ningbo Yongjiang Talent-Introduction Programme(No.2022A-023-C).
文摘The viscosity of refining slags plays a critical role in metallurgical processes.However,obtaining accurate viscosity data remains challenging due to the complexities of high-temperature experiments,often relying on empirical models with limited predictive capabilities.This study focuses on the influence of optical basicity on viscosity in CaO-Al_(2)O_(3)-based refining slags,leveraging machine learning to address data scarcity and improve prediction accuracy.An automated framework for algorithm integration,parameter tuning,and evaluation ranking framework(Auto-APE)is employed to develop customized data-driven models for various slag systems,including CaO-Al_(2)O_(3)-SiO_(2),CaO-Al_(2)O_(3)-CaF_(2),CaO-Al_(2)O_(3)-SiO_(2)-MgO,and CaO-Al_(2)O_(3)-SiO_(2)-MgO-CaF_(2).By incorporating optical basicity as a key feature,the models achieve an average validation error of 8.0%to 15.1%,significantly outperforming traditional empirical models.Additionally,symbolic regression is introduced to rapidly construct domain-specific features,such as optical basicity-like descriptors,offering a potential breakthrough in performance prediction for small datasets.This work highlights the critical role of domain-specific knowledge in understanding and predicting viscosity,providing a robust machine learning-based approach for optimizing refining slag properties.
基金funded by the National Natural Science Foundation of China(Grant No.62441212)the Major Project of the Natural Science Foundation of Inner Mongolia(Grant No.2025ZD008).
文摘Aiming at the problem of insufficient recognition of implicit variants by existing Chinese sensitive text detection methods,this paper proposes the IPKE-MoE framework,which consists of three parts,namely,a sensitive word variant extraction framework,a sensitive word variant knowledge enhancement layer and a mixture-of-experts(MoE)classification layer.First,sensitive word variants are precisely extracted through dynamic iterative prompt templates and the context-aware capabilities of Large Language Models(LLMs).Next,the extracted variants are used to construct a knowledge enhancement layer for sensitive word variants based on RoCBert models.Specifically,after locating variants via n-gram algorithms,variant types are mapped to embedding vectors and fused with original word vectors.Finally,a mixture-of-experts(MoE)classification layer is designed(sensitive word,sentiment,and semantic experts),which decouples the relationship between sensitiveword existence and text toxicity throughmultiple experts.This framework effectively combines the comprehension ability of Large Language Models(LLMs)with the discriminative ability of smaller models.Our two experiments demonstrate that the sensitive word variant extraction framework based on dynamically iterated prompt templates outperforms other baseline prompt templates.TheRoCBert models incorporating the sensitive word variant knowledge enhancement layer and a mixture-of-experts(MoE)classification layer achieve superior classification performance compared to other baselines.
基金funded by the Hunan Provincial Natural Science Foundation of China(Grant No.2025JJ70105)the Hunan Provincial College Students’Innovation and Entrepreneurship Training Program(Project No.S202411342056)The article processing charge(APC)was funded by the Project No.2025JJ70105.
文摘With the widespread use of social media,the propagation of health-related rumors has become a significant public health threat.Existing methods for detecting health rumors predominantly rely on external knowledge or propagation structures,with only a few recent approaches attempting causal inference;however,these have not yet effectively integrated causal discovery with domain-specific knowledge graphs for detecting health rumors.In this study,we found that the combined use of causal discovery and domain-specific knowledge graphs can effectively identify implicit pseudo-causal logic embedded within texts,holding significant potential for health rumor detection.To this end,we propose CKDG—a dual-graph fusion framework based on causal logic and medical knowledge graphs.CKDG constructs a weighted causal graph to capture the implicit causal relationships in the text and introduces a medical knowledge graph to verify semantic consistency,thereby enhancing the ability to identify the misuse of professional terminology and pseudoscientific claims.In experiments conducted on a dataset comprising 8430 health rumors,CKDG achieved an accuracy of 91.28%and an F1 score of 90.38%,representing improvements of 5.11%and 3.29%over the best baseline,respectively.Our results indicate that the integrated use of causal discovery and domainspecific knowledge graphs offers significant advantages for health rumor detection systems.This method not only improves detection performance but also enhances the transparency and credibility of model decisions by tracing causal chains and sources of knowledge conflicts.We anticipate that this work will provide key technological support for the development of trustworthy health-information filtering systems,thereby improving the reliability of public health information on social media.
文摘Knowledge distillation has become a standard technique for compressing large language models into efficient student models,but existing methods often struggle to balance prediction accuracy with explanation quality.Recent approaches such as Distilling Step-by-Step(DSbS)introduce explanation supervision,yet they apply it in a uniform manner that may not fully exploit the different learning dynamics of prediction and explanation.In this work,we propose a task-structured curriculum learning(TSCL)framework that structures training into three sequential phases:(i)prediction-only,to establish stable feature representations;(ii)joint prediction-explanation,to align task outputs with rationale generation;and(iii)explanation-only,to refine the quality of rationales.This design provides a simple but effective modification to DSbS,requiring no architectural changes and adding negligible training cost.We justify the phase scheduling with ablation studies and convergence analysis,showing that an initial prediction-heavy stage followed by a balanced joint phase improves both stability and explanation alignment.Extensive experiments on five datasets(e-SNLI,ANLI,CommonsenseQA,SVAMP,and MedNLI)demonstrate that TSCL consistently outperforms strong baselines,achieving gains of+1.7-2.6 points in accuracy and 0.8-1.2 in ROUGE-L,corresponding to relative error reductions of up to 21%.Beyond lexical metrics,human evaluation and ERASERstyle faithfulness diagnostics confirm that TSCL produces more faithful and informative explanations.Comparative training curves further reveal faster convergence and lower variance across seeds.Efficiency analysis shows less than 3%overhead in wall-clock training time and no additional inference cost,making the approach practical for realworld deployment.This study demonstrates that a simple task-structured curriculum can significantly improve the effectiveness of knowledge distillation.By separating and sequencing objectives,TSCL achieves a better balance between accuracy,stability,and explanation quality.The framework generalizes across domains,including medical NLI,and offers a principled recipe for future applications in multimodal reasoning and reinforcement learning.
基金supported by the following projects:National Natural Science Foundation of China(62461041)Natural Science Foundation of Jiangxi Province China(20242BAB25068).
文摘With the large-scale deployment of the Internet ofThings(IoT)devices,their weak securitymechanisms make them prime targets for malware attacks.Attackers often use Domain Generation Algorithm(DGA)to generate random domain names,hiding the real IP of Command and Control(C&C)servers to build botnets.Due to the randomness and dynamics of DGA,traditional methods struggle to detect them accurately,increasing the difficulty of network defense.This paper proposes a lightweight DGA detection model based on knowledge distillation for resource-constrained IoT environments.Specifically,a teacher model combining CharacterBERT,a bidirectional long short-term memory(BiLSTM)network,and attention mechanism(ATT)is constructed:it extracts character-level semantic features viaCharacterBERT,captures sequence dependencieswith the BiLSTM,and integrates theATT for key feature weighting,formingmulti-granularity feature fusion.An improved knowledge distillation approach transfers the teacher model’s learned knowledge to the simplified DistilBERT student model.Experimental results show the teacher model achieves 98.68%detection accuracy.The student modelmaintains slightly improved accuracy while significantly compressing parameters to approximately 38.4%of the teacher model’s scale,greatly reducing computational overhead for IoT deployment.
文摘Amidst evolving user behavior driven by the development of the internet,enhancing the operational quality of trade publishing knowledge service platforms has become a significant challenge for publishing institutions.To address this issue,this paper employs a combined approach of theoretical analysis and case study,introducing the SICAS(Sense-Interest-Connection-Action-Share)user consumption behavior analysis model and selecting“CITIC Academy”as the case study subject.It systematically examines and summarizes the platform’s operational practices and specific strategies,aiming to offer strategic insights and practical references for the operational improvement and sustainable,high-quality development of trade publishing knowledge service platforms.
文摘Conversational recommender systems(CRSs)focus on refining preferences and providing personalized recommendations through natural language interactions and dialogue history.Large language models(LLMs)have shown outstanding performance across various domains,thereby prompting researchers to investigate their applicability in recommendation systems.However,due to the lack of task-specific knowledge and an inefficient feature extraction process,LLMs still have suboptimal performance in recommendation tasks.Therefore,external knowledge sources,such as knowledge graphs(KGs)and knowledge bases(KBs),are often introduced to address the issue of data sparsity.Compared to KGs,KBs possess higher retrieval efficiency,making them more suitable for scenarios where LLMs serve as recommenders.To this end,we introduce a novel framework integrating LLMs with KBs for enhanced retrieval generation,namely LLMKB.LLMKB initially leverages structured knowledge to create mapping dictionaries,extracting entity-relation information from heterogeneous knowledge to construct KBs.Then,LLMKB achieves the embedding calibration between user information representations and documents in KBs through retrieval model fine-tuning.Finally,LLMKB employs retrievalaugmented generation to produce recommendations based on fused text inputs,followed by post-processing.Experiment results on two public CRS datasets demonstrate the effectiveness of our framework.Our code is publicly available at the link:https://anonymous.4open.science/r/LLMKB-6FD0.
基金supported by the National Natural Science Foundation of China(42476084,62203456,42276199)the Stable Support Project of National Key Laboratory(WDZC 20245250302)the National Key R&D Program of China(2024YFC2813502,2024YFC2813302)。
文摘Bird's-eye-view(BEV)perception is a core technology for autonomous driving systems.However,existing solutions face the dilemma of high costs associated with multimodal methods and limited performance of vision-only approaches.To address this issue,this paper proposes a framework named“a lightweight pure visual BEV perception method based on dual distillation of spatial-temporal knowledge”.This framework innovatively designs a lightweight vision-only student model based on Res Net,which leverages a dual distillation mechanism to learn from a powerful teacher model that integrates temporal information from both image and light detection and ranging(LiDAR)modalities.Specifically,we distill efficient multi-modal feature extraction and spatial fusion capabilities from the BEVFusion model,and distill advanced temporal information fusion and spatiotemporal attention mechanisms from the BEVFormer model.This dual distillation strategy enables the student model to achieve perception performance close to that of multi-modal models without relying on Li DAR.Experimental results on the nu Scenes dataset demonstrate that the proposed model significantly outperforms classical vision-only algorithms,achieves comparable performance to current state-of-the-art vision-only methods on the nu Scenes detection leaderboard in terms of both mean average precision(mAP)and the nu Scenes detection score(NDS)metrics,and exhibits notable advantages in inference computational efficiency.Although the proposed dual-teacher paradigm incurs higher offline training costs compared to single-model approaches,it yields a streamlined and highly efficient student model suitable for resource-constrained real-time deployment.This provides an effective pathway toward low-cost,high-performance autonomous driving perception systems.
基金supported in part by National Natural Science Foundation of China(62441605)。
文摘Large language models(LLMs)have significantly advanced artificial intelligence(AI)by excelling in tasks such as understanding,generation,and reasoning across multiple modalities.Despite these achievements,LLMs have inherent limitations including outdated information,hallucinations,inefficiency,lack of interpretability,and challenges in domain-specific accuracy.To address these issues,this survey explores three promising directions in the post-LLM era:knowledge empowerment,model collaboration,and model co-evolution.First,we examine methods of integrating external knowledge into LLMs to enhance factual accuracy,reasoning capabilities,and interpretability,including incorporating knowledge into training objectives,instruction tuning,retrieval-augmented inference,and knowledge prompting.Second,we discuss model collaboration strategies that leverage the complementary strengths of LLMs and smaller models to improve efficiency and domain-specific performance through techniques such as model merging,functional model collaboration,and knowledge injection.Third,we delve into model co-evolution,in which multiple models collaboratively evolve by sharing knowledge,parameters,and learning strategies to adapt to dynamic environments and tasks,thereby enhancing their adaptability and continual learning.We illustrate how the integration of these techniques advances AI capabilities in science,engineering,and society—particularly in hypothesis development,problem formulation,problem-solving,and interpretability across various domains.We conclude by outlining future pathways for further advancement and applications.
基金Private Higher Education Teaching Quality Management Professional Committee of the Guangdong Higher Education Teaching Management Association,“Research on the Reconstruction of Curriculum System for Visual Communication Major in Private Colleges Driven by Digital Intelligence Technology”(Project No.:GDZLGL25063)Guangzhou College of Commerce“AI+”Pilot Program:“Design Thinking and Innovative Design”(Project No.:2024rgznsdkc09)。
文摘To further improve the level and quality of art design education in colleges and universities and promote the reform and innovation of higher education,this paper conducts research from the perspective of intangible cultural heritage(ICH)knowledge graph empowering education,aiming to break through the current bottlenecks in the integration of ICH and art education and advance the innovative reform of art design education in colleges and universities.Firstly,it analyzes and interprets the significance of ICH knowledge graph empowering college art design education.Then,it elaborates on the challenges faced in this process.Finally,it proposes multiple educational measures,including optimizing graph construction,strengthening literacy education,deepening curriculum integration,and improving the support system,so as to effectively promote the innovative reform of college art design education and enhance the level and quality of talent cultivation in colleges and universities.
基金supported by the National Key R&D Program of China(2018AAA0101502)the Science and Technology Project of SGCC(State Grid Corporation of China):Fundamental Theory of Human-in-the-Loop Hybrid-Augmented Intelligence for Power Grid Dispatch and Control。
文摘Knowledge graphs(KGs)have been widely accepted as powerful tools for modeling the complex relationships between concepts and developing knowledge-based services.In recent years,researchers in the field of power systems have explored KGs to develop intelligent dispatching systems for increasingly large power grids.With multiple power grid dispatching knowledge graphs(PDKGs)constructed by different agencies,the knowledge fusion of different PDKGs is useful for providing more accurate decision supports.To achieve this,entity alignment that aims at connecting different KGs by identifying equivalent entities is a critical step.Existing entity alignment methods cannot integrate useful structural,attribute,and relational information while calculating entities’similarities and are prone to making many-to-one alignments,thus can hardly achieve the best performance.To address these issues,this paper proposes a collective entity alignment model that integrates three kinds of available information and makes collective counterpart assignments.This model proposes a novel knowledge graph attention network(KGAT)to learn the embeddings of entities and relations explicitly and calculates entities’similarities by adaptively incorporating the structural,attribute,and relational similarities.Then,we formulate the counterpart assignment task as an integer programming(IP)problem to obtain one-to-one alignments.We not only conduct experiments on a pair of PDKGs but also evaluate o ur model on three commonly used cross-lingual KGs.Experimental comparisons indicate that our model outperforms other methods and provides an effective tool for the knowledge fusion of PDKGs.
文摘In the international shipping industry, digital intelligence transformation has become essential, with both governments and enterprises actively working to integrate diverse datasets. The domain of maritime and shipping is characterized by a vast array of document types, filled with complex, large-scale, and often chaotic knowledge and relationships. Effectively managing these documents is crucial for developing a Large Language Model (LLM) in the maritime domain, enabling practitioners to access and leverage valuable information. A Knowledge Graph (KG) offers a state-of-the-art solution for enhancing knowledge retrieval, providing more accurate responses and enabling context-aware reasoning. This paper presents a framework for utilizing maritime and shipping documents to construct a knowledge graph using GraphRAG, a hybrid tool combining graph-based retrieval and generation capabilities. The extraction of entities and relationships from these documents and the KG construction process are detailed. Furthermore, the KG is integrated with an LLM to develop a Q&A system, demonstrating that the system significantly improves answer accuracy compared to traditional LLMs. Additionally, the KG construction process is up to 50% faster than conventional LLM-based approaches, underscoring the efficiency of our method. This study provides a promising approach to digital intelligence in shipping, advancing knowledge accessibility and decision-making.
文摘In the context of power generation companies, vast amounts of specialized data and expert knowledge have been accumulated. However, challenges such as data silos and fragmented knowledge hinder the effective utilization of this information. This study proposes a novel framework for intelligent Question-and-Answer (Q&A) systems based on Retrieval-Augmented Generation (RAG) to address these issues. The system efficiently acquires domain-specific knowledge by leveraging external databases, including Relational Databases (RDBs) and graph databases, without additional fine-tuning for Large Language Models (LLMs). Crucially, the framework integrates a Dynamic Knowledge Base Updating Mechanism (DKBUM) and a Weighted Context-Aware Similarity (WCAS) method to enhance retrieval accuracy and mitigate inherent limitations of LLMs, such as hallucinations and lack of specialization. Additionally, the proposed DKBUM dynamically adjusts knowledge weights within the database, ensuring that the most recent and relevant information is utilized, while WCAS refines the alignment between queries and knowledge items by enhanced context understanding. Experimental validation demonstrates that the system can generate timely, accurate, and context-sensitive responses, making it a robust solution for managing complex business logic in specialized industries.
