0.25
0.5
0.75
1.25
1.5
1.75
2
Inducing Implicit Relations from Text using Distantly Supervised Deep Nets
Published on Nov 22, 20182873 Views
In this paper we present Socrates, a deep learning based solutions for Automated Knowledge Base Population. Socrates does not require hand labelled data for domain adaptation. Instead, it exploits a p
Related categories
Chapter list
Inducing Implicit Relations from text using distantly supervised Deep Nets00:00
Outline - 100:23
Extending Knowledge Graphs using Distant Supervision01:53
Binary Contexts03:20
Deep Learning Model04:03
Socrates Architecture for Knowledge Graph Population05:19
Unary Relations05:51
Unary Contexts06:52
Locate single occurrences of entities07:31
Socrates Architecture for Knowledge Graph Population08:03
Gathering Composite Contexts from Title Oriented Documents08:51
Outline - 209:49
Binary Context: evaluation in New York Times - Freebase09:54
Unary Contexts: Evaluation in Common Crawl – DBPedia10:31
Outline - 311:17
ISWC 2017 Challenge11:23
ISWC 2017 Challenge: Data Collection and Methodology 12:00
Attribute Prediction12:27
Attribute Validation12:54
Conclusion and Future work13:15
Inducing Implicit Relations from text using distantly supervised Deep Nets15:03