Analyzing 3D Visualizations of Knowledge Graphs using XAI-Based Routines for Network Analytics and Network Explainability

Graph + AI Summit Industry and Partner Day on October 15, 2021

HubSpot Video

We demonstrate that the high-level network-structure of a Knowledge Graph can be revealed by constructing interactive, real-time 3D visualizations of a Knowledge Graph. We also explain how insights can automatically be extracted from a Knowledge Graph with the use of AI-based routines for network analytics and network explainability (XAI). We illustrate these ideas with two case studies. The first case study, pertaining to life sciences, presents the visualization and analysis of a knowledge graph encoding a metabolic network that has been enriched with multi-omics data. The second case study, which involves the analysis of a Knowledge Graph extracted from a corpus of natural language documents, shows how a visualization of the Knowledge Graph lets us quickly identify groups of related documents and understand what characterizes each of these groups and how they relate to one another.


  • Sagar Indurkhya, Virtualitics