Chaos Orchestra - The Knowledge Graph Podcast   /     #08 - Graph Representation Learning - Guiseppe Futia

Summary

Graph Neural Networks are very effective in dealing with complex network data structures to perform label and link predictions. They can process typological and structural information from social networks to protein pathways. But can they also work with multi-dimensional and dynamic data models of Semantic Graphs? What information loss does one have to consider when it comes to Machine Learning based on ontologies?

Subtitle
Duration
1544
Publishing date
2021-08-05 05:00
Contributors
  Boris Shalumov
author  
Enclosures
https://www.buzzsprout.com/1767483/8973823-08-graph-representation-learning-guiseppe-futia.mp3
audio/mpeg

Shownotes

Graph Neural Networks are very effective in dealing with complex network data structures to perform label and link predictions. They can process typological and structural information from social networks to protein pathways. But can they also work with multi-dimensional and dynamic data models of Semantic Graphs? What information loss does one have to consider when it comes to Machine Learning based on ontologies?