The Semantic Web - ISWC 2023

Name: The Semantic Web - ISWC 2023 | 22nd International Semantic Web Conference, Athens, Greece, November 6-10, 2023, Proceedings, Part I
Brand: Springer
Price: 90.94 EUR
Availability: OnlineOnly

22nd International Semantic Web Conference, Athens, Greece, November 6-10, 2023, Proceedings, Part I

Terry R. Payne Valentina Presutti Guilin Qi María Poveda-Villalón Giorgos Stoilos Laura Hollink Zoi Kaoudi Gong Cheng Juanzi Li(Herausgeber*in)

Springer (Verlag)

Erschienen am 1. November 2023

XLV, 640 Seiten

E-Book

PDF mit Wasserzeichen-DRM

Systemvoraussetzungen

978-3-031-47240-4 (ISBN)

90,94 €inkl. 7% MwSt.

Systemvoraussetzungen

für PDF mit Wasserzeichen-DRM

E-Book Einzellizenz

Als Download verfügbar

Beschreibung

Weitere Details

Weitere Ausgaben

Inhalt

Intro
Preface
Organization
Keynote Speeches and Panels
Knowledge Graphs in the Age of Large Language Models
Semantic Web Research in the Age of Generative Artificial Intelligence
ChatGLM: An Alternative to ChatGPT
Panel on Neuro-Symbolic AI
Contents - Part I
Contents - Part II
Research Track
Link Traversal Query Processing Over Decentralized Environments with Structural Assumptions
1 Introduction
2 Related Work
3 The Solid Ecosystem
3.1 Data Vault
3.2 WebID Profile
3.3 Type Index
4 Approach
4.1 Formal Preliminaries
4.2 Pipeline-Based Link Queue
4.3 Discovery of Data Vault
4.4 Discovery of Type Index
4.5 Implementation
5 Evaluation
5.1 Experimental Design
5.2 Experimental Results
5.3 Discussion
6 Conclusions
Supplemental Material Statement
References
Dense Re-Ranking with Weak Supervision for RDF Dataset Search
1 Introduction
2 Dense Re-Ranking for RDF Dataset Search
2.1 Overview
2.2 Compact Document Representation
2.3 Dense Re-Ranking
3 Coarse Tuning with Weak Supervision
3.1 Overview
3.2 Coarse-Tuning Based on Distant Supervision
3.3 Coarse-Tuning Based on Self-training
4 Evaluation
4.1 Test Collection
4.2 Labeled Data for Coarse-Tuning
4.3 Implementation Details
4.4 Experimental Settings
4.5 Evaluation Metrics
4.6 Evaluation Results
4.7 Case Study
5 Related Work
5.1 Dataset Search
5.2 Dense Ranking
6 Conclusion and Future Work
References
Optimizing SPARQL Queries with SHACL
1 Introduction
2 Preliminaries
3 Optimizations
4 Discussion
5 Conclusion
References
HAEE: Low-Resource Event Detection with Hierarchy-Aware Event Graph Embeddings
1 Introduction
2 Related Work
3 Methodology
3.1 Problem Formulation
3.2 Model Overview
3.3 Event Detection
3.4 Graph Embedding
3.5 Embedding Convolution
3.6 Loss Function
4 Experiments
4.1 Datasets
4.2 Experiments Settings
4.3 Baselines
4.4 Results
5 Analysis
5.1 Cause Representation
5.2 Subevent Representation
5.3 Ablation Studies
6 Conclusions and Future Work
References
Textual Entailment for Effective Triple Validation in Object Prediction
1 Introduction
2 Related Work
3 Object Prediction
3.1 Task Definition
3.2 SATORI: Seek and Entail for Object Prediction
4 Evaluation Setup
4.1 Dataset
4.2 Metrics
4.3 SATORI Configuration
4.4 Baselines
4.5 Pretrained Models and Training Regimes
5 Evaluation Results
5.1 Language Model Prompting and Triple Validation
5.2 Additional Pretraining and Fine-Tuning Scenarios
5.3 Language Models Vs Structured Knowledge Sources and Information Extraction Methods
6 Conclusions
References
ASKRL: An Aligned-Spatial Knowledge Representation Learning Framework for Open-World Knowledge Graph
1 Introduction
2 Related Work
2.1 Closed-World KRL Models
2.2 Open-World KRL Models
2.3 Inductive KRL Models
3 Methodology
3.1 Preliminary
3.2 Framework Overview
3.3 Structured Embedding Layer
3.4 Description Encoding Layer
3.5 Embedding Space Alignment Layer
3.6 Training Optimization
3.7 Three-Stage Optimization Strategy
4 Experiment
4.1 Datasets and Evaluation Metrics
4.2 Baselines
4.3 Implementation Details
4.4 Main Results
4.5 Results of Ablation Study
4.6 Case Study
5 Conclusion
References
Rethinking Uncertainly Missing and Ambiguous Visual Modality in Multi-Modal Entity Alignment
1 Introduction
2 Related Work
3 Method
3.1 Preliminaries
3.2 Multi-modal Knowledge Embedding
3.3 Multi-scale Modality Hybrid
3.4 Circularly Missing Modality Imagination
3.5 Training Details
4 Experiment
4.1 Experiment Setup
4.2 Overall Results
4.3 Details Analysis
5 Conclusion
References
Disentangled Contrastive Learning for Knowledge-Aware Recommender System
1 Introduction
2 Preliminaries
3 Methodology
3.1 Multi-aspect Embeddings Initialization
3.2 Knowledge Graph Disentangling Module
3.3 Interaction Graph Disentangling Module
3.4 Contrastive Learning Module
3.5 Model Prediction
3.6 Multi-task Training
4 Experiment
4.1 Experiment Settings
4.2 Performance Comparison (RQ1)
4.3 Ablation Studies (RQ2)
4.4 Sensitivity Analysis (RQ3)
5 Related Work
5.1 Knowledge-Aware Recommendation
5.2 Disentangled Representation Learning
5.3 Contrastive Learning
6 Conclusion
References
Dependency-Aware Core Column Discovery for Table Understanding
1 Introduction
2 Preliminary
3 Rough Matching and Dependency Scores
3.1 Rough Matching
3.2 Dependency Score
4 DaCo: A Two-Level Iterative Algorithm
4.1 Outer Iteration
4.2 Inner Iteration
5 Experiment
5.1 Experimental Settings
5.2 Overall Performance (RQ1)
5.3 Effects of Unlinkable Mentions (RQ2-1)
5.4 Effects of Row Sampling (RQ2-2)
5.5 Effects of Multiple Core Columns (RQ2-3)
5.6 Effects of Varying Parameter Settings (RQ3)
5.7 Performance on Table Understanding Tasks (RQ4)
6 Related Work
7 Conclusion
References
Spatial Link Prediction with Spatial and Semantic Embeddings
1 Introduction
2 Problem Statement
3 Approach
3.1 Spatial and Semantic Embedding of Entities
3.2 Supervised Spatial Link Prediction Approach
3.3 Unsupervised Spatial Link Prediction Approach
4 Evaluation Setup
4.1 Datasets
4.2 Ground Truth Creation
4.3 Baselines
4.4 Evaluation Metrics
5 Evaluation
5.1 Performance in Transductive Setting
5.2 Performance in Inductive Setting
6 Related Work
7 Conclusion
References
How is Your Knowledge Graph Used: Content-Centric Analysis of SPARQL Query Logs
1 Introduction
2 Related Work
3 Preliminaries
3.1 Query Templates
3.2 Query Log Summarisation Problem
4 Approach
5 Experimentation
5.1 The Dataset
5.2 Methodology of Analysis
5.3 Results
5.4 Discussion
6 Conclusions
References
Iterative Geographic Entity Alignment with Cross-Attention
1 Introduction
2 Problem Statement
3 The IGEA Approach
3.1 Geographic Class Alignment
3.2 Candidate Generation
3.3 Cross-Attention-Based Entity Alignment
3.4 Iterative Geographic Entity Alignment Approach
4 Evaluation Setup
4.1 Datasets
4.2 Ground Truth
4.3 Baselines
4.4 Evaluation Metrics
5 Evaluation
5.1 Entity Alignment Performance
5.2 Ablation Study
5.3 Impact of the Number of Iterations
5.4 Alignment Threshold Tuning
5.5 Manual Assessment of New Links
6 Related Work
7 Conclusion
References
Entity-Relation Distribution-Aware Negative Sampling for Knowledge Graph Embedding
1 Introduction
2 Related Work
2.1 Knowledge Graph Embedding Models
2.2 Negative Sampling Methods
3 Theoretical Analysis
3.1 Optimal Solution for the Negative Sampling Loss Function
3.2 Inaccessible Optimal Solution
3.3 Assigning Varying N(e,r) to Entity-Relation Pair (e,r)
4 Entity-Relation Distribution-Aware Negative Sampling Method (ERDNS)
4.1 From Theory to Practice: Two-Step Expansion
4.2 The Proposed Method
5 Experiments
5.1 Experimental Setting
5.2 Main Results
5.3 Ablation Study
6 Conclusion
References
Negative Sampling with Adaptive Denoising Mixup for Knowledge Graph Embedding
1 Introduction
2 Preliminaries
3 Methodology
3.1 Marginal Pseudo-Negative Triple Estimator (MPNE)
3.2 Adaptive Mixup(AdaMix)
3.3 Traning the KGE Model
4 Experiment
4.1 Experiment Settings
4.2 Experimental Results (Q1)
4.3 Combine with Other NS (Q2)
4.4 Ablation Study (Q3)
4.5 Further Analysis
5 Related Work
6 Conclusion and Future Work
References
Comparison of Knowledge Graph Representations for Consumer Scenarios
1 Introduction
2 Knowledge Representation Models
3 Datasets
4 Knowledge Exploration Scenario
4.1 Experimental Setup
4.2 Results
5 Systematic Querying Scenario
5.1 Experimental Setup
5.2 Results
6 Knowledge Graph Embedding Scenario
6.1 Experimental Setup
6.2 Results
7 Related Work
8 Discussion and Conclusions
8.1 Summary of Results
8.2 Limitations
References
A Comprehensive Study on Knowledge Graph Embedding over Relational Patterns Based on Rule Learning
1 Introduction
2 Related Work
3 Preliminaries and Background
4 Methodology
4.1 Classification of Triples
4.2 Score-Based Patterns Adaptation (SPA)
5 Experiments
5.1 Evaluation Setup
5.2 Quantitative Analysis over Relational Patterns
5.3 Score-Based Patterns Adaptation Results
5.4 Case Study
6 Conclusion and Future Work
References
Compact Encoding of Reified Triples Using HDTr
1 Introduction
2 Reifying RDF Statements
3 Related Work on Efficient Reification Representation
4 HDT
5 Extending HDT for Reified Triples: HDTr
5.1 The HDTr Dictionary
5.2 The HDTr Triples
5.3 Querying HDTr
6 Evaluation
6.1 HDTr vs. HDT
6.2 HDTr vs. Triplestores
7 Conclusions and Future Work
References
Causal Inference-Based Debiasing Framework for Knowledge Graph Completion
1 Introduction
2 Related Work
2.1 Knowledge Graph Completion
2.2 Bias in Knowledge Graph
2.3 Causal Inference
3 Preliminaries
3.1 Notation
3.2 Inference Process of KGC Methods
4 Methodology
4.1 Bias Analysis
4.2 Causal Analysis
4.3 In-Depth Bias Mitigation
4.4 In-Breadth Bias Mitigation
5 Experiment
5.1 Datasets
5.2 Evaluation Metrics
5.3 Baselines
5.4 Hyperparameters
5.5 Main Results
6 Analysis
6.1 Fine-Grained Analysis
6.2 Effect of In-Depth Bias Mitigation
7 Conclusion and Future Work
References
Can ChatGPT Replace Traditional KBQA Models? An In-Depth Analysis of the Question Answering Performance of the GPT LLM Family
1 Introduction
2 Related Work
2.1 Large Language Models and Prompting
2.2 Evaluation of the Large Language Model
2.3 Black-Box Testing of the NLP Model
3 Evaluation Framework
3.1 Feature-Driven Unified Question Labeling
3.2 Answer Evaluation
3.3 CheckList Testing
4 Experiments
4.1 Datasets
4.2 Comparative Models
4.3 Overall Results
4.4 Multilingual KBQA Results
4.5 Feature Tags Based Results
4.6 CheckList Results
5 Conclusion
References
Mapping and Cleaning Open Commonsense Knowledge Bases with Generative Translation
1 Introduction
2 Previous Work
2.1 Commonsense Knowledge Bases
2.2 From Open KBs to Closed KBs
2.3 Existing Systems
3 Problem Formulation
4 Methodology
4.1 Creating Weakly-Labelled Training Data
4.2 Generative Translation Mapping
5 Experiment Setup
5.1 Evaluation
5.2 Baselines
5.3 Implementation
6 Results and Discussion
6.1 Comparison with Baselines
6.2 What Is the Best Alignment Method?
6.3 What Are the Properties of Our Model?
6.4 How Does GenT Compare with Direct LM Generation Methods?
7 Conclusion
References
Integrating Knowledge Graph Embeddings and Pre-trained Language Models in Hypercomplex Spaces
1 Introduction
2 Related Work
2.1 Structural Knowledge Graph Embedding Models
2.2 Text-Enhanced Knowledge Graph Embedding Models
3 Preliminaries
4 Our Method
4.1 Entity, Relation, and Query Representation
4.2 Dimension Adjustment
4.3 Triple Plausibility Score and Training
5 Experiments
5.1 Experimental Setup
5.2 Link Prediction Results and Analysis
5.3 Effect of Textual Information in Entity Representation
5.4 Contribution of Individual Sentences
6 Conclusion
References
LLMs4OL: Large Language Models for Ontology Learning
1 Introduction
2 Related Work
3 The LLMs4OL Task Paradigm
4 LLMs4OL - Three Ontology Learning Tasks Evaluations
4.1 Evaluation Datasets - Ontological Knowledge Sources
4.2 Evaluation Models-Large Language Models (LLMs)
4.3 Evaluations
5 Error Analysis and Limitations
6 Conclusions and Future Directions
References
Biomedical Knowledge Graph Embeddings with Negative Statements
1 Introduction
2 Related Work
2.1 Exploring Negative Statements
2.2 Knowledge Graph Embeddings
2.3 Walk-Based Embeddings
2.4 Tailoring Knowledge Graph Embeddings
3 Methods
3.1 Problem Formulation
3.2 Overview
3.3 Creation of the RDF Graph
3.4 Random Walk Generation with Negative Statements
3.5 Neural Language Models
3.6 Final Representations
4 Experiments
4.1 Data
4.2 Results and Discussion
5 Conclusion
References
Knowledge Graph Enhanced Language Models for Sentiment Analysis
1 Introduction
2 Related Work
2.1 Incorporating External Knowledge for NLP
2.2 Incorporating External Knowledge for Sentiment Analysis
3 Methodology
3.1 Task Definition and Model Overview
3.2 Commonsense-Senti Knowledge Graph Construction
3.3 CS-Knowledge Graph Initialization
3.4 Textual Pre-embedding
3.5 Fusion Module
3.6 Inference and Learning
4 Experiment
4.1 Datasets and External Knowledge
4.2 Comparison Methods
4.3 Implementation Details
4.4 Overall Results
4.5 Ablation Results
4.6 Advantages and Drawbacks of Combining KGs and LMs
5 Conclusion
References
TemporalFC: A Temporal Fact Checking Approach over Knowledge Graphs
1 Introduction
2 Preliminaries
3 Related Work
3.1 Fact Checking
3.2 Time-Point Prediction
4 Methodology
4.1 TKGE Model
4.2 Fact Checking
4.3 Time-Point Prediction
5 Experiments
5.1 Datasets
5.2 Evaluation Metrics
5.3 Setup Details and Reproducibility
5.4 Competing Approaches
6 Results and Discussion
6.1 Fact Checking
6.2 Time-Point Prediction
7 Conclusion
References
Assessing the Generalization Capabilities of Neural Machine Translation Models for SPARQL Query Generation
1 Introduction
2 Related Work
3 Methodology
3.1 Task and Data
3.2 NMT Models
3.3 Evaluation of Generalization Capabilities
4 Experiments
4.1 Data
4.2 Models Training and Evaluation
5 Results
5.1 Original Results
5.2 Unknown URIs Split
5.3 Unknown Templates Split
6 Discussion
7 Conclusion and Future Work
References
Linking Tabular Columns to Unseen Ontologies
1 Introduction
2 Approach
2.1 Encoding the Query Column and the Headers
2.2 Encoding the Target Type
2.3 Putting it All Together
3 Dataset Construction
3.1 Building Datasets Using WikiTables
3.2 Building Datasets Using BioDivTab
3.3 Tough Tables Dataset
4 Experiments and Analysis
4.1 Results on WikiTables and DBpedia
4.2 Results on BioDivTab and 2T
5 Related Work
6 Conclusion
7 Limitations
A Appendix
A.1 Model Predictions
References
Neural Multi-hop Logical Query Answering with Concept-Level Answers
1 Introduction
2 Related Work
3 Methodology
3.1 Incorporating Concepts
3.2 Representing Concepts
3.3 Representing Instances and Queries
3.4 Operating on Concepts
3.5 Optimization and Inference
4 Experiments
4.1 Experimental Settings
4.2 Experimental Results
5 Discussion
6 Conclusion
References
ForecastTKGQuestions: A Benchmark for Temporal Question Answering and Forecasting over Temporal Knowledge Graphs
1 Introduction
2 Preliminaries and Related Work
3 ForecastTKGQuestions
3.1 Temporal Knowledge Base
3.2 Question Categorization and Generation
4 ForecastTKGQA
4.1 TKG Forecasting Model
4.2 QA Model
5 Experiments
5.1 Experimental Setting
5.2 Main Results
5.3 Human Vs. Machine on FRQs
5.4 Performance over FRQs with Different Reasoning Types
5.5 Answerability of ForecastTKGQuestions
5.6 Challenges of Forecasting TKGQA over Forecast TKG Questions
5.7 Study of Data Efficiency
6 Justification of Task Validity from Two Perspectives
7 Conclusion
References
SORBET: A Siamese Network for Ontology Embeddings Using a Distance-Based Regression Loss and BERT
1 Introduction
2 Related Work
2.1 Ontology and Knowledge Graph Embeddings
2.2 Tasks Related to Ontology Embeddings
3 Methodology
3.1 Architecture
3.2 Data Sampling
3.3 Regression Distance-Based Loss
4 Experiments
4.1 Datasets
4.2 Model
4.3 Experiments on the Ontology Alignment Task
4.4 Experiments on the Subsumption Task
4.5 Ablation Study
5 Discussion
5.1 Ontological Representation
5.2 Superposed Ontologies
5.3 Limitations of Our Approach
6 Conclusion
References
Visualizing Mappings Between Pairwise Ontologies - An Empirical Study of Matrix and Linked Indented List in Their User Support During Class Mapping Creation and Evaluation
1 Introduction
2 Related Work
3 Experimental Setup
3.1 Tasks and Visualizations
3.2 Participants, Protocol, and Data Collection
4 Results
5 Conclusions and Future Work
References
FeaBI: A Feature Selection-Based Framework for Interpreting KG Embeddings
1 Introduction
2 Preliminaries
3 Embedding-Guided Feature Construction
3.1 Feature Construction
3.2 Feature Selection
4 Experiments
4.1 Experimental Setup
4.2 Experimental Results
5 Related Work
6 Conclusion
References
CapsKG: Enabling Continual Knowledge Integration in Language Models for Automatic Knowledge Graph Completion
1 Introduction
2 Related Work
3 Task Definition: Link Prediction
4 Methodology
4.1 Base Model: LM for Link Prediction
4.2 Continual Learning for Link Prediction
5 Experiments
5.1 Datasets
5.2 Experimental Setup
5.3 Link Prediction Evaluation
5.4 Low Resource Evaluation
5.5 Topic Evaluation
6 Conclusion
References
Author Index

Systemvoraussetzungen

Als PDF speichern Als Link merken