The Semantic Web : 21st International Conference, ESWC 2024, Hersonissos, Crete, Greece, May 26-30, 2024, Proceedings, Part I

• Autore: Meroño Peñuela Albert
• Pubblicazione: Cham : , : Springer, , 2024; ©2024
• Edizione: 1st ed.
• Descrizione fisica: 1 online resource (0 pages)
• Altri autori: DimouAnastasia TroncyRaphaël HartigOlaf AcostaMaribel AlamMehwish PaulheimHeiko LisenaPasquale
• Nota di contenuto: Intro -- Preface -- Organization -- Contents - Part I -- Contents - Part II -- Research -- Do Similar Entities Have Similar Embeddings? -- 1 Introduction -- 2 Related Work -- 3 Approach -- 3.1 Towards a Graph-Based Notion of Similarity -- 3.2 Embedding vs. Graph: A Different Notion of Similarity? -- 4 Experiments -- 4.1 Knowledge Graph Embedding Models -- 4.2 Datasets -- 5 Results and Discussion -- 5.1 Different Notions of Similarity (RQ1) -- 5.2 Correlation Between Rank-Based Metrics and RBO (RQ2) -- 5.3 Analyzing Predicate Importance (RQ3) -- 6 Conclusion and Outlook -- References -- Treat Different Negatives Differently: Enriching Loss Functions with Domain and Range Constraints for Link Prediction -- 1 Introduction -- 2 Related Work -- 2.1 Semantic-Enhanced Approaches -- 2.2 Loss Functions for the Link Prediction Task -- 3 Signature-Driven Loss Functions -- 4 Experimental Setting -- 4.1 Evaluation Metrics -- 4.2 Datasets and Models -- 4.3 Implementation Details -- 5 Results -- 5.1 Global Performance -- 5.2 Ablation Study -- 6 Discussion -- 6.1 Treating Different Negatives Differently (RQ1) -- 6.2 Impact of Signature-Driven Losses on Performance (RQ2) -- 6.3 On the Evaluation of KGEM Predictions for LP -- 7 Conclusion -- A Modified Versions of LSBCEL and LSPLL -- B Bucket Analysis -- References -- QAGCN: Answering Multi-relation Questions via Single-Step Implicit Reasoning over Knowledge Graphs -- 1 Introduction -- 2 Related Work -- 3 QAGCN - A Question-Aware GCN-Based Question Answering Model -- 3.1 Task Definition -- 3.2 The Question-Aware GCN -- 4 Experiments -- 4.1 Effectiveness Evaluation -- 4.2 Comparison with SOTA Reasoning-Based Method - NSM -- 4.3 Efficiency Evaluation -- 4.4 Ablation Study -- 4.5 Case Study with Embedding Visualization -- 5 Conclusion and Outlook -- References. ; Leveraging Pre-trained Language Models for Time Interval Prediction in Text-Enhanced Temporal Knowledge Graphs -- 1 Introduction -- 2 Preliminaries -- 3 The Framework -- 4 Experiments -- 4.1 The Datasets -- 4.2 Evaluation Metrics -- 4.3 Experimental Setup -- 4.4 Transductive Time Interval Prediction -- 4.5 Inductive Time Interval Prediction -- 4.6 Fine-Grained Analysis -- 5 Related Work -- 6 Conclusion -- 7 Appendix -- 7.1 Effect of Number of Negative Samples -- 7.2 Evaluation Terminology -- 7.3 Variance Analysis -- References -- A Language Model Based Framework for New Concept Placement in Ontologies -- 1 Introduction -- 2 Related Work -- 2.1 Ontology Concept Placement -- 2.2 Pre-trained Language Models for Ontology Concept Placement -- 3 Problem Statement -- 4 Methodology -- 4.1 Edge Search: Searching Seed Concepts or Edges -- 4.2 Edge Formation and Enrichment -- 4.3 Edge Selection -- 5 Experiments -- 5.1 Data Construction -- 5.2 Metrics -- 5.3 Experimental Settings and Baseline Methods -- 5.4 Results -- 5.5 Ablation Studies -- 5.6 Conclusion and Future Studies -- References -- Low-Dimensional Hyperbolic Knowledge Graph Embedding for Better Extrapolation to Under-Represented Data -- 1 Introduction -- 2 Related Work -- 3 Problem Setup -- 3.1 Knowledge Graph Embedding (KGE) -- 3.2 Riemannian Geometry -- 3.3 Extrapolation -- 3.4 Composition Patterns -- 4 Method: HolmE -- 4.1 HolmE: Holomorphic KG Embedding -- 4.2 Model Analysis -- 5 Experiments -- 5.1 Overall Experiment Design -- 5.2 Extrapolation to Unseen Triples -- 5.3 Extrapolation to Under-Represented Entities -- 5.4 Extrapolation to Under-Represented Relations -- 6 Conclusion -- A Result Details -- References -- SC-Block: Supervised Contrastive Blocking Within Entity Resolution Pipelines -- 1 Introduction -- 2 WDC-Block: A Large Product Blocking Benchmark. ; 3 SC-Block: Supervised Contrastive Blocking -- 4 Blocking-Only Evaluation -- 4.1 Baseline Blocking Methods -- 4.2 Implementation -- 4.3 Results for Fixed K -- 4.4 Results for 99.5% Recall on Validation Set -- 5 Evaluation Within Entity Resolution Pipelines -- 5.1 F1 Score and Runtime -- 5.2 Impact of Training Time -- 6 Related Work -- 7 Conclusion -- A Appendix - Additional Experiments -- References -- Navigating Ontology Development with Large Language Models -- 1 Introduction -- 2 Related Work -- 2.1 Ontology Engineering -- 2.2 Ontology Learning and Generation -- 2.3 LLMs for Code Generation -- 3 Methodology -- 3.1 Large Language Models -- 3.2 Prompting Methods -- 3.3 Experimental Setup -- 4 Experimental Results -- 4.1 Initial Experiment Results -- 4.2 Main Experiment Results -- 5 Discussion -- 6 Conclusions and Future Work -- References -- ESLM: Improving Entity Summarization by Leveraging Language Models -- 1 Introduction -- 2 Related Work -- 2.1 Entity Summarization -- 2.2 Contextual Language Models -- 3 Approach -- 3.1 Problem Statement -- 3.2 ESLM Model -- 3.3 ESLM Model Enrichment -- 3.4 Entity Summarization -- 4 Experimental Setup -- 4.1 Baselines -- 4.2 Datasets -- 4.3 Experimental Settings -- 5 Results and Analysis -- 5.1 Comparison with State-of-the-Art Approaches -- 5.2 Example Findings -- 5.3 Ablation Study -- 5.4 Computational Requirements and Efficiency of ESLM -- 6 Conclusion and Future Work -- References -- Explanation of Link Predictions on Knowledge Graphs via Levelwise Filtering and Graph Summarization -- 1 Introduction -- 2 Fundamentals -- 2.1 Knowledge Graphs and Embedding Models -- 2.2 Quotient Graph -- 3 Related Works -- 4 The Proposed Approach -- 4.1 Kelpie -- 4.2 Injecting Semantics in the Pre-Filter -- 4.3 Injecting Semantics in the Explanation Builder -- 5 Experimental Evaluation -- 5.1 Experimental Setting. ; 5.2 Quantitative Evaluation -- 5.3 Qualitative Evaluation -- 6 Conclusions -- A Appendix: Repair of Inconsistent and Unsatisfiable Ontologies -- B Appendix: Hyper-parameters -- References -- Large Language Models for Scientific Question Answering: An Extensive Analysis of the SciQA Benchmark -- 1 Introduction -- 2 Related Work -- 3 Background -- 3.1 Dataset -- 3.2 Approaches -- 4 Methodology -- 4.1 Experiments -- 4.2 Experimental Setup -- 5 Results and Discussion -- 6 Error Analysis -- 7 Limitations -- 8 Conclusions -- A Appendix - Examples -- References -- Efficient Evaluation of Conjunctive Regular Path Queries Using Multi-way Joins -- 1 Introduction -- 2 Preliminaries -- 2.1 RDF and SPARQL -- 2.2 Worst-Case Optimal Multi-way Joins -- 2.3 GenericJoinCRPQ -- 3 Related Work -- 4 Evaluating C2RPQs with Multi-way Joins -- 4.1 Evaluation of RPQs -- 4.2 Evaluation of C2RPQs -- 4.3 Query Planning and Optimizations -- 5 Experimental Results -- 5.1 Systems and Experimental Setup -- 5.2 Datasets and Queries -- 5.3 Evaluation of Execution Strategies -- 5.4 Comparison with Existing Solutions -- 6 Conclusion and Future Work -- References -- Can Contrastive Learning Refine Embeddings -- 1 Introduction -- 2 Preliminaries and Problem Definition -- 2.1 Contrastive Learning -- 2.2 Problem Statement -- 3 Method -- 3.1 Contrastive Learning Limitation -- 3.2 SimSkip Details -- 3.3 Data Augmentation -- 3.4 Theoretical Proof -- 4 Experiments -- 4.1 Experimental Setting -- 4.2 SimSkip for Federated Knowledge Graph Embedding -- 4.3 SimSkip for Image Embedding -- 4.4 SimSkip for Node embedding learned by Supervised Learning -- 4.5 SimSkip for Transformer-based Text Embedding -- 4.6 Ablation Study -- 5 Related Work -- 5.1 Representation Learning -- 5.2 Contrastive Learning -- 6 Conclusion -- References -- In-Use. ; Automation of Electronic Invoice Validation Using Knowledge Graph Technologies -- 1 Introduction -- 2 Preliminaries -- 3 Our Approach -- 3.1 EDIFACT Ontology -- 3.2 The EDIFACT-VAL Tool -- 4 Evaluation -- 4.1 Experimental Set up -- 4.2 Results of the Ontology Evaluation -- 4.3 Completeness of the Invoice KG Generated by EDIFACT-Val -- 4.4 Results of Invoice Validation with EDIFACT-Val -- 4.5 Runtime Performance of EDIFACT-Val -- 4.6 In-Use Evaluation -- 5 Related Work -- 6 Conclusion and Future Work -- References -- Towards Cyber Mapping the German Financial System with Knowledge Graphs -- 1 Introduction -- 2 Related Work -- 3 Knowledge Graph Construction -- 3.1 Cyber Mapping Ontology -- 3.2 Structured Data: Financial Domain -- 3.3 Unstructured Data: Fund Prospectus -- 3.4 Provenance Information -- 4 Knowledge Graph Application -- 5 User Study -- 5.1 Setup -- 5.2 Results -- 6 Conclusion and Outlook -- References -- Integrating Domain Knowledge for Enhanced Concept Model Explainability in Plant Disease Classification -- 1 Introduction -- 2 Use Cases -- 3 Related Work -- 4 Methods -- 4.1 Ontology Development (Ontology Based Explanation) -- 4.2 Synthetic Concepts Images Generation -- 4.3 Testing with Concept Activation Vectors (TCAV) -- 5 Experiments and Results -- 5.1 Dataset and Trained Model -- 5.2 Experimental Setup -- 5.3 Findings and Analyses -- 6 Conclusion -- References -- Generative Expression Constrained Knowledge-Based Decoding for Open Data -- 1 Introduction -- 2 Related Work -- 3 Application -- 3.1 Generative Models -- 4 Evaluation -- 4.1 Quantitative Results -- 4.2 Qualitative User Testing -- 5 Impact and Deployment -- 6 Conclusions and Future Work -- A S-expression functions considered in GECKO -- References -- OntoEditor: Real-Time Collaboration via Distributed Version Control for Ontology Development -- 1 Introduction -- 2 Motivation. ; 3 Related Work.
• Titolo autorizzato: The Semantic Web
• ISBN: 3-031-60626-4
• Formato: Materiale a stampa
• Livello bibliografico: Monografia
• Lingua di pubblicazione: Inglese
• Record Nr.: 9910861091703321
• Lo trovi qui: Univ. Federico II
• Serie: Lecture Notes in Computer Science Series