0.25
0.5
0.75
1.25
1.5
1.75
2
Matching Web Tables with Knowledge Base Entities: From Entity Lookups to Entity Embeddings
Published on Nov 28, 2017968 Views
Web tables constitute valuable sources of information for various applications, ranging from Web search to Knowledge Base (KB) augmentation. An underlying common requirement is to annotate the rows of
Related categories
Chapter list
Matching Web Tables with Knowledge Base Entities: From Entity Lookups to Entity Embeddings00:00
Outline00:13
What is a Knowledge Graph (KG)?00:38
FactBase: a Knowledge Graph API for AI applications00:48
Web table annotation01:11
Model and assumptions01:29
Main challenges02:12
Related work - lookups03:21
Related work - entity disambiguation based on embeddings [ZSG - ESWC ‘16]04:12
Related work – ontology matching tools05:27
Contributions05:52
Our approaches for annotating entities: FactBase lookup06:23
Our approaches for annotating entities: Entity embeddings method08:29
Entity disambiguation – the graph of candidates08:44
Entity disambiguation – the selected candidates09:43
Our approaches for annotating entities: Ontology matching method10:02
A hybrid approach10:34
Experimental evaluation - Datasets11:06
Wikipedia gold standard11:31
Issues11:43
Experimental results with T2D gold standard11:56
Experimental results with Limaye gold standard12:43
Experimental results with our Wikipedia gold standard13:26
Lessons Learned14:21
Conclusion14:51