Automating the discovery of linking candidates / by Michael Röder ; 1. Reviewer: Prof. Dr. Axel-Cyrille Ngonga Ngomo, 2. Reviewer: Prof. Dr. Elena Demidova. Paderborn, 2023
Inhalt
- Titlepage
- 1 Introduction
- 2 Preliminaries
- 3 Benchmarking Linked Data Systems
- 3.1 Related Work
- 3.2 Benchmarking with Linked Data
- 3.3 Requirements
- 3.4 Platform Architecture
- 3.4.1 Overview
- 3.4.2 Platform Components
- 3.4.3 Benchmark Components
- 3.4.4 Benchmarked System Components
- 3.4.5 Benchmark Workflow
- 3.5 Mimicking real-world RDF Graphs
- 3.5.1 Graph Analysis
- 3.5.2 Learning Graph Invariants
- 3.5.3 Initial Graph Generation
- 3.5.4 Graph Amendment
- 3.5.5 Graph Completion
- 3.6 Evaluation
- 3.7 Application
- 3.8 Limitations and Future Work
- 3.9 Conclusion
- 4 Crawling the Web of Data
- 4.1 Related work
- 4.2 Web of Data Crawler
- 4.3 Crawling Benchmark
- 4.4 Evaluation
- 4.4.1 Benchmarked Crawlers
- 4.4.2 Data Web Crawling
- 4.4.3 Efficiency Evaluation
- 4.4.4 Robots Exclusion Protocol Check
- 4.4.5 Evaluation with Lemming Graphs
- 4.5 Discussion
- 4.6 Application
- 4.7 Conclusion
- 5 A Topic Model for the Data Web
- 5.1 Related Work
- 5.2 LODCat
- 5.2.1 Topic Inference
- 5.2.2 Model Evaluation
- 5.2.3 Topic Labeling
- 5.2.4 RDF Dataset Transformation
- 5.2.5 Topic Assignment
- 5.3 Topic Evaluation
- 5.3.1 Framework of Coherence Measures
- 5.3.2 Evaluation Setup
- 5.3.3 Results and Discussion
- 5.3.4 Runtimes
- 5.3.5 Application in LODCat
- 5.4 Evaluation of LODCat
- 5.5 Conclusion
- 6 Dataset Search for Linking
- 6.1 Related Work
- 6.2 Our Approach
- 6.2.1 Metadata Extraction
- 6.2.2 Document Generation
- 6.2.3 Topic Model Inference
- 6.2.4 Similarity Calculation
- 6.3 Benchmarking Dataset Linking Recommendation Systems
- 6.4 Evaluation
- 6.5 Conclusion
- 7 Summary
- Bibliography
- List of Abbreviations
- List of RDF Namespaces
- List of Symbols
- A Appendix